framework,version,device,op_name,kernel_source,gemm_dtype,m,n,k,latency
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,16384,32.79494900173611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,51200,103.02298990885417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,51200,153.83025444878473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,12288,33.28144666883681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,16384,47.945048014322914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,16384,84.00292290581596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,10240,20.94581773546007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,12288,35.82193332248264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,12288,63.165554470486114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,8192,19.766021728515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,10240,30.209442138671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,10240,52.645229763454864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,8192,23.85421074761285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,8192,41.238427056206596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,7168,19.37016126844618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,6144,15.929514567057291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,7168,21.120391845703125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,7168,36.29675971137153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,5120,10.444456312391493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,6144,18.281063503689236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,6144,31.30929904513889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,5120,15.253175523546007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,4096,10.695811801486544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,5120,26.125345865885418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,3584,7.379601372612847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,4096,20.942379421657986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,51200,261.76982964409723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,3584,10.624276055230036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,3584,18.26454332139757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,3072,6.295359293619792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,2560,5.690645429823134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,3072,8.616637335883246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,3072,12.448812696668837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,2560,7.09981706407335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,2048,4.811537848578559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,2560,6.570557488335504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,2048,5.371266682942708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,2048,5.019547356499566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,1536,4.144273969862196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,1024,2.738247977362739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,1536,4.050614251030816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,1536,3.9420683119032116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,768,2.0397858089870877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,1024,2.974540498521593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,768,2.3679991828070746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,768,2.2918968200683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,512,1.2454142040676541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,512,1.8738488091362848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,256,0.9230133692423502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,512,1.7949769761827257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,256,1.3653920491536458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,256,1.6625199847751195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,128,0.7509422302246094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,128,1.5855467054578993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,64,0.6371084319220649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,64,1.6129955715603297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,4096,12.054655287000868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,65536,32,0.6578782399495443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,65536,32,2.087715572781033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,1024,2.8227600521511502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,65536,128,1.0277093251546223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,65536,103.71180555555556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,51200,80.38123236762154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,51200,124.18705240885417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,65536,162.60344780815973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,65536,266.5621744791667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,16384,34.83024427625868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,12288,24.201705932617188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,16384,38.35045030381944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,51200,207.65208604600696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,16384,66.35237630208333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,10240,19.094256930881077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,12288,28.856387668185764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,8192,16.01046413845486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,12288,49.732818603515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,10240,24.314517550998264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,7168,14.221551683213976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,10240,41.58453708224827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,8192,19.470650566948784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,8192,32.56159125434028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,7168,17.112157185872395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,6144,13.022421095106337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,7168,28.673170301649307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,5120,10.63198937310113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,6144,14.72023179796007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,6144,24.768142700195312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,5120,12.291766696506075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,4096,8.255335489908854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,3584,6.9572703043619795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,5120,20.644939846462673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,4096,9.757828606499567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,4096,16.54354010687934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,3072,6.203102111816406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,3584,8.40484873453776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,2560,4.970036400689019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,3584,14.36924574110243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,3072,6.915730794270833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,3072,10.08148447672526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,2560,5.227915445963542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,2560,5.6823243035210504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,2048,4.084475623236762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,2048,4.286892361111112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,1536,3.1199120415581594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,2048,4.165487077501085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,1024,1.8860221438937719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,1536,3.234333250257704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,1536,3.1041280958387585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,1024,2.378798166910807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,768,1.3172026740180123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,1024,2.2758089701334634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,768,1.888167911105686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,512,1.1328390969170463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,768,1.8474710252549913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,256,0.7629573080274793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,512,1.4943795733981664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,512,1.4385795593261719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,256,1.3070080015394423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,128,0.5869377983940972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,256,1.1036880281236436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,64,0.49332533942328555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,128,1.241106669108073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,51200,32,0.5126781993442112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,64,1.2654231389363606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,51200,32,1.6386524836222331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,65536,34.4730224609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,51200,128,0.8164044486151801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,65536,53.795732286241325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,51200,25.872733222113713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,16384,10.454059176974827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,16384,23.351677788628475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,65536,94.76078287760417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,51200,41.57453409830729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,12288,6.567969852023655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,16384,12.774795532226562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,12288,9.58017306857639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,10240,6.265640682644314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,12288,17.61041090223524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,51200,73.42511664496529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,10240,8.07340325249566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,8192,5.041266547309028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,10240,14.734859890407987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,7168,4.3697793748643665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,8192,6.509230295817058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,8192,11.57641347249349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,6144,3.762567308213976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,7168,5.659183078342014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,5120,3.084178712632921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,7168,9.890775044759115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,6144,5.978347354465061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,6144,4.861509535047743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,4096,2.512593799167209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,5120,4.130321926540799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,5120,3.994831085205078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,4096,3.1875875261094837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,3584,1.8228711022271051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,4096,3.191724353366428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,3584,2.7681297726101346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,3072,1.818746566772461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,3584,2.812845230102539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,2560,1.3547200096978083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,3072,2.3658356136745877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,2560,1.9880489773220484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,3072,2.441361745198568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,2048,1.0397058063083224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,2560,2.0557706620958114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,2048,1.6141928566826715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,1536,0.8206008805168999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,2048,1.6957084867689345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,1024,0.6054648823208278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,1536,1.2586417728000217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,1024,0.9153280258178711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,1536,1.3213724560207791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,768,0.47459734810723203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,1024,0.9703582127888998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,768,0.7142657703823514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,512,0.37492799758911133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,768,0.775310198465983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,256,0.32857778337266713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,512,0.5945768886142307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,512,0.5575644705030653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,256,0.45541069242689347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,128,0.19342133733961317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,256,0.4542186525132921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,64,0.1660675605138143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,16384,128,0.3044711218939887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,16384,32,0.16811645030975342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,64,0.4152533213297526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,128,0.41427821583218044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,16384,32,0.47427643669976133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,51200,20.10928683810764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,65536,25.38796827528212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,65536,45.617024739583336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,16384,6.741515689425999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,51200,35.31915283203125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,16384,18.432654486762154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,16384,10.780126783582899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,65536,74.57395426432292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,12288,5.445000966389974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,51200,57.892985026041664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,12288,8.122152540418837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,10240,5.415264129638672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,12288,13.868374294704862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,8192,3.46632448832194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,10240,11.372066921657987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,10240,6.776060316297744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,7168,3.183095932006836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,8192,6.3857209947374125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,8192,5.387784745958116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,6144,2.761930677625868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,7168,4.836675431993273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,7168,4.702875349256727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,6144,3.972661336263021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,5120,2.427439159817166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,6144,4.0112792121039496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,4096,1.7517671585083008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,5120,3.23598141140408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,5120,3.348012500339084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,4096,2.5571511586507163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,3584,1.673998196919759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,4096,2.7097305721706815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,3584,2.251183191935221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,3072,1.4574835035536025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,3584,2.3861520555284286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,2560,1.1836062537299263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,3072,1.9362373352050781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,3072,2.066572401258681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,2048,0.7764044337802463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,2560,1.6317164103190105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,2560,1.7534595065646703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,2048,1.3245439529418945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,1536,0.6550809012518989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,2048,1.439825799730089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,1024,0.47611554463704425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,1536,1.0306062698364258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,1536,1.1165795856051974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,1024,0.7469484541151258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,768,0.3476693365308974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,1024,0.8525324397616916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,512,0.24785955746968588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,768,0.5786026848687066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,256,0.17912622292836508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,512,0.4490017890930176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,768,0.6468142403496636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,512,0.49210664961073136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,128,0.14629244804382324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,256,0.35765600204467773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,64,0.12563021977742514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,128,0.31881509886847603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,128,0.24274489614698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,12288,32,0.13362488481733534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,64,0.3153395652770996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,12288,32,0.3396000067392985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,12288,256,0.333734220928616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,51200,16.167056613498264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,65536,21.210316975911457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,65536,41.14742702907986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,16384,5.512880113389756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,51200,32.33830430772569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,65536,64.08737521701389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,16384,15.938978407118055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,16384,9.854318406846788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,51200,50.01848347981771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,12288,4.194194793701172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,10240,3.6282098558213978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,12288,7.429514567057292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,8192,3.146620432535807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,10240,7.472141689724392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,12288,11.77410634358724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,10240,6.1609751383463545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,7168,2.9355494181315103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,8192,4.784865061442058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,8192,4.7991820441351996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,7168,4.098110198974609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,6144,2.2724541558159723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,7168,4.222532484266493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,5120,2.0383421579996743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,6144,3.4520276387532554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,6144,3.6390469868977866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,4096,1.4667413499620225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,5120,2.870272954305013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,4096,2.280104955037435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,5120,3.0538586510552297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,3584,1.3406062655978734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,4096,2.4650613996717667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,3584,2.003634770711263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,3072,1.2002524269951713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,3584,2.176908493041992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,3072,1.7270373238457575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,2560,1.0313235388861763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,3072,1.8919662899441188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,2048,0.8478177918328179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,2560,1.4537715911865234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,2560,1.5937253104315863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,1536,0.5421084298027886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,2048,1.1820817523532443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,2048,1.3212240007188585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,1024,0.38314400778876406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,1536,0.9157920413547092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,1536,1.0243493186102974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,768,0.2886879973941379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,1024,0.6607155799865723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,1024,0.7994053098890516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,768,0.5141448974609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,512,0.21378844314151338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,512,0.3935671117570665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,768,0.5839591026306152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,256,0.15140444702572292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,512,0.43687645594278973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,128,0.12243021859063043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,256,0.2926933235592312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,256,0.31016890207926434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,128,0.27231645584106445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,64,0.10416533549626668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,10240,128,0.21116267310248482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,10240,32,0.10896266831292047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,64,0.26843465699089897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,10240,32,0.2820551130506727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,51200,14.023464626736112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,65536,18.344046698676216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,65536,37.013020833333336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,16384,4.467455122205946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,51200,29.119696723090275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,65536,50.81610785590278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,51200,37.511688232421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,12288,3.6046117146809897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,16384,11.967586093478731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,16384,8.884957207573786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,10240,3.3704444037543406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,12288,7.013533274332683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,12288,6.645621405707465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,8192,2.6156275007459855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,10240,5.266081068250868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,10240,5.477560679117839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,7168,2.107422298855252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,8192,4.024560716417101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,8192,4.367971632215712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,6144,1.809722688462999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,7168,3.5148328145345054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,7168,3.834197150336372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,5120,1.4415067036946614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,6144,3.0005946689181857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,6144,3.297815958658854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,4096,1.2126728693644206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,5120,2.5034444597032333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,5120,2.7583128611246743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,3584,1.0217786365085177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,4096,1.9958888159857855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,4096,2.233194563123915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,3584,1.7558604346381292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,3072,0.8135991096496582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,3584,1.9685306549072266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,2560,0.7364266713460287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,3072,1.514712863498264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,3072,1.7047164705064561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,2560,1.2735332912868924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,2048,0.5558044645521376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,2560,1.4491831461588542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,1536,0.4531982209947374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,2048,1.0369990666707356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,1024,0.31129421128167045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,2048,1.1803963979085286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,1536,0.8038960032992892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,768,0.229165342119005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,1536,0.9145466486612955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,1024,0.5748568640814887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,1024,0.6571715672810873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,512,0.1766959958606296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,768,0.44547287623087567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,256,0.12082489331563313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,768,0.5200319819980198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,512,0.33884533246358234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,512,0.3829600016276042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,128,0.10066577461030747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,256,0.2618951002756755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,64,0.08360444174872504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,128,0.1797377798292372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,128,0.22577865918477377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,8192,256,0.25304799609714085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,8192,32,0.08608088890711467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,64,0.22001866499582926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,8192,32,0.22357956568400064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,51200,11.679803636338976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,65536,17.497146606445312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,65536,35.11512586805556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,16384,3.8244285583496094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,65536,41.67841932508681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,51200,31.762953016493057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,51200,27.295995076497395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,12288,2.7212693956163196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,16384,10.06570519341363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,16384,8.36177486843533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,12288,5.945631239149306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,10240,2.475858688354492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,12288,6.257616678873698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,8192,2.21181509229872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,10240,4.761177062988281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,10240,5.170024871826172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,8192,3.6970719231499563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,7168,1.9076675838894313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,6144,1.8229484558105469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,8192,4.14281251695421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,7168,3.258870230780708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,7168,3.623327891031901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,5120,1.3467804590861003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,6144,2.7817760043674045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,4096,1.0588933096991646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,6144,3.1243839263916016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,5120,2.326364517211914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,5120,2.6166239844428167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,3584,0.9461457994249133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,4096,1.8597297668457031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,4096,2.1186534033881292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,3072,0.7424355612860786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,3584,1.6341981887817383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,3584,1.8685857984754775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,2560,0.6215679910447862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,3072,1.4096417956882055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,2048,0.49524535073174375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,3072,1.6182275348239477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,2560,1.1842390696207683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,1536,0.4009919961293538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,2560,1.3621457417805989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,2048,0.9666177961561414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,2048,1.1160675684611003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,1024,0.27220087581210667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,1536,0.7855057716369629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,1536,0.8615919748942057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,768,0.20925866232977974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,1024,0.5329635408189561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,512,0.14710488584306505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,1024,0.6191493140326606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,768,0.411740435494317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,256,0.10655288563834296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,768,0.48672978083292645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,512,0.3099599944220649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,128,0.08719199895858765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,512,0.35776178042093915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,256,0.23219201299879286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,64,0.07274311118655734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,128,0.20137066311306426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,7168,32,0.07611111137602064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,7168,128,0.16495022508833143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,64,0.1948444445927938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,32,0.19733777311113146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,7168,256,0.23352977964613172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,51200,11.541619194878471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,65536,14.060351053873697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,65536,33.24895562065972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,65536,35.00196668836806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,16384,3.3608534071180554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,51200,26.87600877549913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,51200,25.730266994900173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,12288,2.4518765343560114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,16384,8.288120693630642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,12288,5.313281589084201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,16384,7.855762057834202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,10240,2.2467447916666665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,8192,1.8198346032036676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,12288,5.869578467475043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,10240,4.326500362820096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,10240,4.876726362440321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,8192,3.4006631639268665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,8192,3.898128933376736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,7168,1.6959120432535808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,6144,1.2749439875284831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,7168,3.0040452745225696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,6144,2.578195571899414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,7168,3.4240798950195312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,5120,1.048472934299045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,4096,0.9085653093126086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,6144,2.9447644551595054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,5120,2.157644483778212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,5120,2.472139570448134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,3584,0.700351979997423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,4096,1.7176301744249132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,4096,1.9961013793945312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,3072,0.6278808911641439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,3584,1.5080817540486653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,2560,0.508096906873915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,3584,1.761329756842719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,3072,1.301836437649197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,3072,1.5308586756388347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,2048,0.4233200020260281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,2560,1.1220329072740343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,1536,0.32706578572591144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,2048,0.9169457753499349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,2560,1.299324459499783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,2048,1.0574764675564237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,1024,0.2188755538728502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,1536,0.7302391264173719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,1024,0.4905075497097439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,1536,0.8113706906636556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,768,0.1828391154607137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,512,0.13258666462368435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,1024,0.5813413196139866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,768,0.3792444335089789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,768,0.45523113674587673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,256,0.09247910976409912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,512,0.31439110967848033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,128,0.08152977625528972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,512,0.3314470979902479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,256,0.21041066116756865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,256,0.2120666636361016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,64,0.06483111116621229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,128,0.17880622545878092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,64,0.1703733338250054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,6144,32,0.06853777832455106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,6144,32,0.17063466707865396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,65536,11.004099527994791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,51200,8.85499996609158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,65536,29.465718587239582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,65536,31.252610948350693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,16384,2.8859920501708984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,51200,22.90673828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,16384,6.7649171617296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,12288,2.242916531032986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,6144,128,0.14914845095740423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,12288,4.787552727593316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,16384,7.37628173828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,10240,1.7095919715033636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,12288,5.50688976711697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,8192,1.4447733561197917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,10240,3.9468650817871094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,51200,24.232133653428818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,7168,1.3102987077501085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,10240,4.589141421847873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,8192,3.1318738725450306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,6144,1.123761812845866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,8192,3.6830728318956165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,7168,2.760102166069878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,5120,0.9028231302897135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,7168,3.229530758327908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,6144,2.368751949734158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,4096,0.6471946504380968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,6144,2.783032947116428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,5120,1.9843618604871962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,5120,2.3308472103542752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,3584,0.5705164273579916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,4096,1.5815991295708551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,4096,1.8851227230495875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,3072,0.5209475623236762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,3584,1.3897573682996962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,3584,1.6561120351155598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,2560,0.4475279914008246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,3072,1.2006461885240343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,3072,1.4362417856852214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,2048,0.3437813387976752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,2560,1.0095964007907443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,2560,1.207765367296007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,1536,0.28168710072835285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,2048,0.8190631336636014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,1024,0.1780088875028822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,2048,0.9872293472290039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,1536,0.6318551169501411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,768,0.1503102249569363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,1536,0.7585902214050293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,1024,0.44835731718275285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,512,0.10016800297631158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,1024,0.5402248700459799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,768,0.3459155559539795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,768,0.42295998997158474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,256,0.07797244522306654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,512,0.25522756576538086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,128,0.06425244278377958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,512,0.30525067117479115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,256,0.19209599494934082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,256,0.1839831140306261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,64,0.05446488989724053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,128,0.15433155165778265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,5120,32,0.05712711148791843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,64,0.1448835531870524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,5120,128,0.13364799817403158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,5120,32,0.14303378264109293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,51200,6.464102427164714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,65536,9.513794793023004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,65536,25.269756740993923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,51200,20.418197631835938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,16384,2.2735652923583984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,65536,29.076914469401043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,12288,1.7131581836276586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,51200,22.7765875922309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,16384,5.980901506212022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,10240,1.3201324674818251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,16384,6.881295098198785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,12288,4.316372341579862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,8192,1.027081807454427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,12288,5.150013393825955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,10240,3.6002638075086804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,10240,4.293982187906901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,7168,0.8857377900017632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,8192,2.8612533145480685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,6144,0.7489457660251193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,8192,3.449615902370877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,7168,2.51620970831977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,5120,0.658807118733724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,7168,3.0323297712537975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,6144,2.1624213324652777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,4096,0.5491920047336155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,6144,2.6058943006727433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,5120,1.8058452606201172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,5120,2.1870924631754556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,3584,0.46681955125596786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,4096,1.4441600375705297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,4096,1.7627084520128038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,3072,0.4046008851793077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,3584,1.2668533325195312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,2560,0.33660888671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,3584,1.553287082248264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,3072,1.0919555028279622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,2048,0.28204798698425293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,3072,1.341294182671441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,2560,0.9283012813991971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,1536,0.21198666095733643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,2560,1.131614261203342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,2048,0.7459191216362847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,1024,0.14051377773284912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,2048,0.9192995495266385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,1536,0.5753031306796604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,768,0.11324089103274876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,1024,0.4055368900299072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,1536,0.708530690934923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,512,0.0856284432941013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,768,0.3128257857428657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,1024,0.4988275633917914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,256,0.062112887700398765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,768,0.38983111911349827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,512,0.2791706720987956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,256,0.15903645091586643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,512,0.22858222325642905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,128,0.05182844400405884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,64,0.04500888784726461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,256,0.17158399687872994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,4096,32,0.04793511165512932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,128,0.1305733389324612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,64,0.12058311038547093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,4096,128,0.11787554952833389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,4096,32,0.11735910839504665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,51200,6.3622283935546875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,65536,7.89168463812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,65536,24.573688083224827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,51200,19.12160407172309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,16384,1.9168194664849176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,65536,28.25928243001302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,51200,22.087461683485245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,12288,1.5867591434054906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,16384,5.635449727376302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,10240,1.2802115546332467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,16384,6.670607248942058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,12288,4.0949219597710504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,8192,1.1436257892184787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,12288,4.997856987847222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,10240,3.417941411336263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,7168,0.965402709113227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,10240,4.170840793185763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,8192,2.7212763892279734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,8192,3.3503430684407554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,6144,0.7010115517510308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,7168,2.3933804829915366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,5120,0.6761484675937228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,7168,2.9395919375949435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,6144,2.05765872531467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,6144,2.524926291571723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,4096,0.49826664394802517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,5120,1.7247431013319228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,3584,0.42235199610392254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,5120,2.118828455607096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,4096,1.3757235209147136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,3072,0.37892889976501465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,4096,1.7096142239040797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,3584,1.2063706715901692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,2560,0.30449599689907497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,3584,1.5028390884399414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,3072,1.0396596060858834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,2048,0.2681244479285346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,3072,1.2995999654134114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,2560,0.903781361050076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,1536,0.18695999516381157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,2560,1.093358251783583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,2048,0.709436469607883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,1024,0.13009244865841335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,2048,0.8894142574734158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,1536,0.5459875530666775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,768,0.0975582202275594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,1024,0.3844257725609674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,1536,0.6838044590420194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,512,0.07675111293792725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,768,0.2964133421579997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,1024,0.4815484682718913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,256,0.05519288778305054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,768,0.37451555993821883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,512,0.21408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,128,0.04640977912478977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,512,0.26459911134507924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,256,0.16231732898288304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,64,0.03901244534386529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,128,0.11865954928927952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,256,0.14684177769554987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3584,32,0.04026133484310574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,64,0.10781688822640313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3584,32,0.1046488881111145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3584,128,0.10944444603390163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,65536,6.879968855116103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,51200,5.210015190972222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,65536,22.001944647894963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,16384,1.651643541124132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,51200,17.725189208984375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,65536,27.33700391981337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,16384,5.327791849772136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,12288,1.2626444498697917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,16384,6.426384819878472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,51200,21.356862386067707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,10240,1.0913315878974068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,12288,3.8920940823025174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,8192,0.8081066873338488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,12288,4.82078128390842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,10240,3.2411679161919484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,7168,0.7066613303290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,10240,4.031482696533203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,8192,2.584954579671224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,6144,0.5933626492818197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,7168,2.2743235694037542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,8192,3.2329004075792103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,5120,0.5242435667249892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,6144,1.950927946302626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,7168,2.839704089694553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,4096,0.41852445072597927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,5120,1.6272800233629015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,6144,2.4425474802652993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,3584,0.3614319960276286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,4096,1.3060471216837566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,5120,2.042881859673394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,3072,0.3295066621568468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,4096,1.646566179063585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,3584,1.1460773679945204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,3584,1.4506426917182074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,2560,0.2698711024390327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,3072,0.9865475760565864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,2048,0.2106168932384915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,2560,0.8300915294223361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,3072,1.251164436340332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,2048,0.6730497678120931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,2560,1.0542471143934462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,1536,0.16200621922810873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,1024,0.1156275537278917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,1536,0.5170284377204047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,1024,0.36271733707851833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,768,0.09186400307549371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,1536,0.6581359969245063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,1024,0.4602515432569716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,768,0.27975111537509495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,512,0.06502755482991536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,768,0.35915644963582355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,256,0.047784000635147095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,512,0.20023643970489502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,512,0.2514417701297336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,128,0.04016888803905911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,256,0.15222756067911783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,2048,0.8568408754136828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,128,0.10668088992436726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,3072,128,0.10149421956804063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,32,0.037992000579833984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,64,0.09487733576032852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,32,0.09103822045856053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,3072,256,0.13385066721174452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,3072,64,0.03594133257865906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,51200,4.31151368882921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,65536,5.46857410007053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,65536,21.242611355251736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,51200,16.61335923936632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,16384,1.2828612857394748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,51200,20.622346666124134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,16384,4.989889780680339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,12288,1.0325440300835502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,65536,26.39611307779948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,10240,0.8688213560316297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,16384,6.191437191433376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,12288,3.6935581631130643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,8192,0.6854391098022461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,10240,3.0683990054660373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,12288,4.65244886610243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,7168,0.6330133544074165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,10240,3.8869840833875866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,8192,2.4511146545410156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,6144,0.5111164516872829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,7168,2.1538978152804904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,8192,3.1246115366617837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,5120,0.40298933453030056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,7168,2.7371874915228953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,6144,1.8486186133490667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,4096,0.32551643583509654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,5120,1.5465066697862413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,6144,2.3529226515028214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,3584,0.2959333260854085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,5120,1.971011479695638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,4096,1.2362915674845378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,3072,0.2668622334798177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,3584,1.085609753926595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,4096,1.5895866817898219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,2560,0.2000959979163276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,3072,0.9388631184895834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,3584,1.3950320349799261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,3072,1.2048479715983074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,2560,0.7880604532029895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,2048,0.1730337805218167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,2560,1.0132373174031575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,1536,0.13423555427127415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,2048,0.6370613310072157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,2048,0.8223857879638672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,1536,0.48800357182820636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,1024,0.09223200215233697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,1024,0.3416737715403239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,768,0.07377155621846516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,1536,0.6314417521158854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,1024,0.43979644775390625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,512,0.05484266744719612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,768,0.2630737887488471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,512,0.18651555644141304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,768,0.34192265404595273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,256,0.04035733474625482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,128,0.03398133317629496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,512,0.23889244927300346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,256,0.12106488810645209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,128,0.09445955355962117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,64,0.028892444239722356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2560,32,0.02985244327121311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,128,0.09477688868840535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,64,0.08293510807885064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2560,32,0.07756266991297404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2560,256,0.14174755414326987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,51200,3.627953847249349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,65536,4.633356306287977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,65536,19.554855346679688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,51200,15.336593627929688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,16384,1.0288062625461154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,16384,4.677548302544488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,51200,19.916222466362846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,12288,0.764978673723009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,65536,25.463443332248264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,10240,0.7335759798685709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,16384,5.969821506076389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,12288,3.4729732937282987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,8192,0.5336115625169542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,10240,2.892319997151693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,12288,4.48504638671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,7168,0.49484888712565106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,10240,3.7480443318684897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,8192,2.3166720072428384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,6144,0.4311777750651042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,8192,3.0041351318359375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,7168,2.6381759643554688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,5120,0.3792275587717692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,6144,1.7463386323716905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,7168,2.030214309692383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,4096,0.2727217674255371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,6144,2.265397389729818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,5120,1.4567866855197484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,4096,1.1651635699801974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,3584,0.2307751178741455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,5120,1.8986640506320531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,3072,0.20781866709391275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,3584,1.0236666997273762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,4096,1.5266631444295247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,2560,0.17761688762240938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,3584,1.342900488111708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,3072,0.8841110865275065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,2048,0.14452355437808567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,2560,0.7422551049126519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,3072,1.1572649214002821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,1536,0.1149208943049113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,2560,0.9739849302503797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,2048,0.6010639932420518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,1024,0.07627111011081271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,2048,0.7886942227681478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,1536,0.4589671028984918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,768,0.06300089094373915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,1536,0.6046462059020996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,1024,0.32059910562303334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,768,0.2460249000125461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,512,0.0440942214594947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,1024,0.4199937714470758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,256,0.03257688879966736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,768,0.32639911439683705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,512,0.171524445215861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,256,0.10802133215798272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,128,0.02751733362674713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,256,0.1321048869027032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,512,0.22584356202019584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,64,0.024163555767801072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,2048,32,0.02519911030928294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,2048,128,0.08613333437177871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,128,0.08283378018273248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,32,0.06458222203784518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,2048,64,0.07089066505432129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,51200,2.632354736328125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,65536,3.6476427714029946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,51200,13.952349344889322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,65536,17.786453247070312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,16384,0.8658222092522515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,16384,4.356793721516927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,12288,0.6092141999138726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,51200,19.29793972439236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,16384,5.771990034315322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,65536,24.700453016493057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,10240,0.5236799981858995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,12288,3.262306637234158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,8192,0.4186657799614801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,10240,2.7138631608751087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,12288,4.339404635959201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,7168,0.3700595431857639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,8192,2.1728941599527993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,10240,3.6221385531955295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,6144,0.3069137732187907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,7168,1.9097190433078344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,8192,2.902067608303494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,5120,0.26701421207851833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,6144,1.6382702721489801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,7168,2.5449350145128036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,4096,0.21646489037407768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,5120,1.367575963338216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,6144,2.190848880343967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,4096,1.0957821740044487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,3584,0.1901315583123101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,5120,1.8323234981960719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,3072,0.16775111357371011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,4096,1.4725555843777125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,3584,0.9626942740546333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,2560,0.13784444332122803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,3072,0.8301653332180448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,3584,1.2945217556423612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,2048,0.11127732859717475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,2560,0.6969884236653646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,3072,1.1153910954793294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,1536,0.08692266543706258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,2560,0.9372995164659289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,2048,0.5639200210571289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,1024,0.06120622158050537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,1536,0.430331548055013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,2048,0.7585760222540961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,768,0.04768622252676222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,1536,0.580584896935357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,1024,0.29903022448221844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,1024,0.4021573331620958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,768,0.22875910335116914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,768,0.3112773365444607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,256,0.026710222164789837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,512,0.15744088755713567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,512,0.2127751111984253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,128,0.022113778524928625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,256,0.09514577521218194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,256,0.12170310815175374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,64,0.019526221685939364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,128,0.0705839991569519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1536,128,0.07810755570729573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,32,0.02035466664367252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,64,0.058471997578938804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1536,32,0.051933334933386914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1536,512,0.03561866614553664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,51200,1.8514578077528212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,65536,2.637634701199002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,51200,12.80784437391493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,16384,0.5803306897481283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,65536,16.348234388563366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,16384,4.065433078342013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,12288,0.42965067757500547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,51200,18.59269036187066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,16384,5.542606353759766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,10240,0.3671395513746474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,12288,3.0492880079481335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,65536,23.801593356662327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,8192,0.3050853411356608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,10240,2.537777794731988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,12288,4.166073693169488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,7168,0.2666773266262478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,8192,2.035389370388455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,10240,3.4818187289767795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,6144,0.22572710778978136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,7168,1.7833902570936415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,8192,2.7880071004231772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,5120,0.17742755677964953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,7168,2.447697745429145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,6144,1.5338186687893336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,4096,0.14795733822716609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,5120,1.278982268439399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,6144,2.100252363416884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,3584,0.13345244195726183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,4096,1.028791109720866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,5120,1.7583849165174696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,3072,0.11441155274709065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,4096,1.4117111629909938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,3072,0.77764802508884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,3584,1.2408711115519206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,2560,0.09264266490936279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,3072,1.0679733488294814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,3584,0.9022924635145398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,2048,0.08208800024456449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,2560,0.6529644330342611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,1536,0.059345775180392794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,2048,0.5263768831888834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,2560,0.8967652850680881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,1024,0.04300355580117968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,2048,0.7244826422797309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,1536,0.40115732616848415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,1536,0.553243531121148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,768,0.034669333034091525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,768,0.2110382186041938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,1024,0.38226932949490017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,512,0.02383111086156633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,768,0.293628454208374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,512,0.14237244923909506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,1024,0.2763004567888048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,256,0.018626666731304593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,512,0.19893599881066215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,128,0.015478221906556023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,256,0.08250488837560017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,256,0.11176533169216579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,128,0.05862222115198771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,64,0.014588443769348992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,1024,32,0.01498311095767551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,64,0.04531733194986979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,1024,32,0.03993689020474752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,65536,1.8635084364149306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,51200,1.6273102230495875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,65536,15.729900783962675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,51200,12.333278232150606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,16384,0.43324889077080625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,16384,3.926979488796658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,65536,23.435150146484375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,12288,0.35907912254333496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,16384,5.451326158311631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,12288,2.9456986321343313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,10240,0.2619680033789741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,12288,4.096307542588976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,10240,2.4578240712483725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,8192,0.25068267186482746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,8192,1.9688631693522136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,10240,3.4187128278944225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,7168,0.19700354999966094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,51200,18.294757419162327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,8192,2.7423617045084634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,7168,1.724287986755371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,6144,0.1850506729549832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,7168,2.4077377319335938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,5120,0.15187822447882757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,6144,1.5110773510403102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,6144,2.079824023776584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,4096,0.11603021621704102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,5120,1.2605244318644206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,5120,1.7251520156860352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,3584,0.10217244757546319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,4096,0.9942639668782552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,4096,1.3853901757134333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,3072,0.09195911222034031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,3584,0.8716897964477539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,3584,1.2180577384101021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,3072,0.7499279975891113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,2560,0.07588266664081149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,3072,1.0482702255249023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,2048,0.06126133600870768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,2560,0.6289804246690538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,2560,0.8795982466803657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,2048,0.5079831017388238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,1536,0.048566222190856934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,2048,0.7092195616828071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,1024,0.03507288959291246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,1536,0.3853004508548313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,1536,0.5417208671569824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,1024,0.26402399275037974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,768,0.027847111225128174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,1024,0.37260622448391384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,512,0.01867022282547421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,768,0.20109244187672934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,512,0.13350133101145426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,512,0.19272977775997588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,256,0.014166222678290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,256,0.07627555396821764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,256,0.1068097750345866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,128,0.01239377756913503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,128,0.0522017776966095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,128,0.06619289186265734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,64,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,768,32,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,64,0.03846399982770284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,768,32,0.03163911236657037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,768,768,0.2866835594177246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,65536,1.1921218236287434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,1024,128,0.07062133153279622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,51200,0.9738924238416883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,51200,11.843721177842882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,16384,0.28243645032246906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,65536,15.122426350911459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,16384,3.7819014655219187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,12288,0.2143688864178128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,51200,17.902128431532116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,12288,2.8384568956163196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,16384,5.323396470811632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,10240,0.19465243816375732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,10240,2.369924545288086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,65536,22.92691718207465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,8192,0.16739022731781006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,10240,3.335137685139974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,12288,3.998308393690321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,7168,0.1517368952433268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,8192,1.8992409176296656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,8192,2.676190270317925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,6144,0.12533422311147055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,7168,1.6634577645195856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,5120,0.10494933525721233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,6144,1.4290044572618272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,5120,1.1926497353447807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,7168,2.346177842881944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,4096,0.08407111300362481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,6144,2.012685351901584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,3584,0.07504622141520183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,4096,0.9590791066487631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,4096,1.3510221905178492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,3584,0.8400444454616971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,5120,1.683068487379286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,3072,0.06481244166692098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,2560,0.05061955584420098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,3072,0.7222648726569282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,3584,1.1870462629530165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,2560,0.6052035755581325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,2048,0.04467555549409655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,3072,1.02165773179796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,1536,0.03559199968973795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,2048,0.4878871175977919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,2560,0.8553848796420627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,1536,0.3702880011664496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,1024,0.024217777782016333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,2048,0.6915804545084635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,768,0.018015111486117046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,1536,0.5255315568712022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,1024,0.2526008817884657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,512,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,1024,0.3605564435323079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,768,0.19243021806081137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,256,0.010427555276287926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,512,0.12717866897583008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,768,0.276493337419298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,512,0.18488710456424287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,128,0.009684444301658207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,256,0.10198399755689834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,128,0.043776889642079674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,64,0.009116444322797988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,512,128,0.06292444467544556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,512,32,0.009993777506881291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,64,0.031192888816197712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,32,0.02572622232966953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,65536,0.7239724265204536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,512,256,0.06716533501942952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,51200,0.5795191129048666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,51200,11.41833241780599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,16384,0.1983893314997355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,65536,14.585830688476562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,16384,3.652716318766276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,12288,0.147962662908766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,51200,17.618094550238716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,12288,2.7423858642578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,16384,5.222966088189019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,10240,0.11968444453345405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,10240,2.285691579182943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,65536,22.52985127766927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,8192,0.09984799888398911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,12288,3.926328870985243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,7168,0.08530933327145046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,10240,3.2773439619276257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,8192,1.8306658003065321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,6144,0.07700888978110419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,7168,1.6046959559122722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,8192,2.6230896843804254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,5120,0.06435110833909777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,7168,2.301483578152127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,6144,1.377211570739746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,4096,0.05183822247717115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,5120,1.1502168443467882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,6144,1.9757946862114801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,3584,0.04678844412167867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,4096,0.9230409198337131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,5120,1.6513937844170465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,3072,0.040898667441474065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,4096,1.3244568506876628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,3584,0.8100026448567709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,2560,0.03522133496072557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,3584,1.161900414360894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,3072,0.696749316321479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,2048,0.029600001043743555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,2560,0.5832631323072646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,3072,0.9999981986151801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,1536,0.02342222299840715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,2048,0.4694853358798557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,2560,0.8379999796549479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,1024,0.013297778036859302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,2048,0.6765217781066895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,1536,0.35510844654507107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,768,0.010821333361996545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,1024,0.2425759898291694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,768,0.18415733178456625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,1536,0.515441788567437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,1024,0.35213955243428546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,512,0.008752889103359645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,768,0.2684560086992052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,512,0.17758666144476998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,256,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,512,0.12084000640445286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,128,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,256,0.06188266807132297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,256,0.09646399815877278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,64,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,256,128,0.056631108125050865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,256,32,0.007985778152942657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,64,0.02548444436656104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,128,0.03569066524505615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,256,32,0.020641777250501845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,65536,0.6617591116163466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,51200,0.5141057968139648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,51200,11.204366895887587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,16384,0.17124711142645943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,65536,14.3407474093967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,16384,3.5884073045518665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,12288,0.13103289074367946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,12288,2.6953555213080516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,51200,17.444193522135418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,10240,0.10916799969143337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,10240,2.2500932481553817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,12288,3.8922623528374563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,8192,0.09353066815270318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,10240,3.2459839714898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,65536,22.34015401204427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,8192,1.8019271426730687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,7168,0.07728266716003418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,7168,1.5789804458618164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,6144,0.0692746639251709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,8192,2.601482603285048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,7168,2.281498591105143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,5120,0.05762044588724772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,6144,1.3557910919189453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,5120,1.13282044728597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,4096,0.047819554805755615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,6144,1.957580354478624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,4096,0.9091350767347547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,3584,0.0421066681543986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,5120,1.6358648935953777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,16384,5.183942159016927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,4096,1.3137022654215496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,3584,0.7972559928894043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,3072,0.03750933210055033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,2560,0.0322728885544671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,3584,1.151249779595269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,3072,0.6851199997795953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,2048,0.026800000005298193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,2560,0.5740844408671061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,3072,0.9919342464870877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,1536,0.017616000440385606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,2560,0.8298168712192111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,2048,0.4613813294304742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,1024,0.008648888932334052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,1536,0.34964977370368105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,1024,0.23722754584418404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,768,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,1536,0.5082240104675293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,1024,0.3458071019914415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,512,0.006525333142942852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,768,0.17963288889990914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,768,0.2637297842237684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,512,0.11770666970147027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,512,0.17365422513749865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,256,0.05942933426962959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,128,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,256,0.09424444701936509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,128,0.032607999112870954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,64,0.004927111168702443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,2048,0.6699760225084094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,32,0.005165333135260476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,64,0.022663111488024395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,128,256,0.0053093334039052325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,128,32,0.019381332728597853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,65536,0.6350684695773655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,51200,0.4953475528293186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,16384,0.17616000440385607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,16384,3.5773468017578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,12288,0.12513777944776747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,12288,2.6876097785102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,10240,0.10473777850468953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,10240,2.243367936876085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,8192,0.08580178022384644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,51200,11.166696336534288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,7168,0.07469599776797824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,8192,1.797690709431966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,6144,0.06560444169574313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,7168,1.5746498107910156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,5120,0.055656890074412026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,6144,1.351668463812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,4096,0.044774221049414746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,65536,14.282769097222221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,3584,0.039827555418014526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,5120,1.1283644570244682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,3072,0.03475022315979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,4096,0.906217787000868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,2560,0.029562665356530085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,3584,0.7943493525187174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,3072,0.6834275457594129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,2048,0.02456444501876831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,2560,0.5717351171705458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,1536,0.013167111410035027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,2048,0.4597075780232747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,1024,0.008394666843944127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,1024,0.2354977660708957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,1536,0.3479146692487929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,768,0.007168889045715332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,512,0.006106666806671355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,512,0.11623199780782063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,256,0.005251555393139522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,768,0.17910310957166883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,128,0.004635555462704765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,256,0.05782310830222236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,64,0.004274666723277834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,128,0.03127911024623447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,64,0.021532444490326777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,64,32,0.004660444541109933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,64,32,0.018237334158685472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32768,128,128,0.055324445168177284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,65536,0.6772986518012153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,16384,0.16772354973687065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,51200,0.5306586689419216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,12288,0.12256533569759792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,65536,14.28153822157118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,10240,0.10273510879940456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,12288,2.6865687900119357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,16384,3.5766135321723094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,8192,0.08760888708962335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,7168,0.07331822315851848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,10240,2.240872065226237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,7168,1.5740346908569336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,6144,0.0649706655078464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,8192,1.7972961001926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,5120,0.05377866824467977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,4096,0.0447982218530443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,4096,0.9062159856160482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,3584,0.04010666741265191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,5120,1.1286239624023438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,6144,1.3515911102294922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,3072,0.0347306662135654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,2560,0.03022577696376377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,3584,0.7980604701571994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,3072,0.6851101981268989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,2048,0.02490933405028449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,1536,0.009377777576446533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,2560,0.5731271107991537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,1024,0.007339555356237624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,2048,0.46225420633951825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,1536,0.34938934114244247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,768,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,512,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,1024,0.23834400706821016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,256,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,768,0.1799164480633206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,512,0.11754843923780654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,128,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,64,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,128,0.030905776553683813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32768,32,32,0.0034328889515664843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,256,0.05860533316930135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,64,0.02201866606871287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,32,0.01718577742576599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32768,32,51200,11.159089830186632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,16384,20.279456244574654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,51200,51.9258287217882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,16384,42.00882297092014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,16384,22.620798746744793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,12288,15.796554565429688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,12288,16.909128824869793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,12288,31.426249186197918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,10240,10.409131368001303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,10240,14.57017347547743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,8192,8.350551181369358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,51200,131.59833441840277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,10240,26.111467149522568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,51200,72.32219780815971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,7168,7.989735921223958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,8192,11.269564310709635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,6144,7.2252146402994795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,7168,9.685221354166666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,8192,20.622645060221355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,7168,17.609805636935764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,5120,5.775484297010634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,5120,6.701991187201606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,6144,10.042872958713108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,6144,8.002049763997396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,4096,4.60721672905816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,5120,6.114946577284072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,4096,5.138711293538411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,4096,4.688297695583768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,3584,3.683582305908203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,3584,4.421248965793185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,3072,3.4028222825792103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,3584,4.094082726372613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,3072,3.7706256442599826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,2560,2.8261627621120877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,3072,3.545248879326714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,2560,3.1838302612304688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,2560,2.9605759514702688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,2048,2.5759440528021917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,2048,2.563647164238824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,1536,1.6576311323377821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,2048,2.458943896823459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,1536,2.0089662339952254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,1024,1.2321280373467338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,1536,1.894958284166124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,1024,1.498928917778863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,768,0.9168542226155599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,1024,1.4093564351399739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,768,1.171372413635254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,512,0.6926915380689832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,768,1.1458142598470051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,512,0.9573556052313911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,256,0.6171937518649632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,512,0.9028284284803602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,128,0.3903244336446126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,256,0.693610668182373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,256,0.8550524181789823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,64,0.32241066296895343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,128,0.8042995664808484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,65536,32,0.3327573405371772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,64,0.8308320045471191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,65536,32,0.9507644441392687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,65536,128,0.525218645731608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,51200,40.73904079861111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,65536,51.37828233506945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,51200,57.430874294704864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,16384,15.970074123806425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,65536,133.68442111545139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,65536,77.15927463107639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,51200,103.65303548177083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,12288,11.019912719726562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,16384,17.68017408582899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,10240,9.29602305094401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,16384,33.21963840060764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,12288,13.31660630967882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,12288,24.84195285373264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,8192,7.318393283420139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,10240,11.46140882703993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,7168,6.586434682210286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,10240,20.721594916449654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,8192,8.805199517144098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,8192,16.289510091145832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,6144,5.667261335584853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,7168,7.612557305230035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,7168,13.893858167860243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,6144,6.443586561414931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,6144,8.116228739420572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,5120,4.646213531494141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,4096,3.4647920396592884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,4096,4.100919935438368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,5120,5.349519941541884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,5120,4.887024773491754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,3584,3.0455449422200522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,4096,3.8183449639214406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,3584,3.544648064507378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,3584,3.326506720648872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,3072,2.6593138376871743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,3072,3.025379604763455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,2560,2.338959164089627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,3072,2.841419643825955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,2560,2.552266650729709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,2048,1.8594159020317926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,2560,2.3759742312961154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,2048,2.0545323689778647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,1536,1.6329563988579645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,2048,1.961037317911784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,1536,1.6106088426378038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,1024,1.0115991168551974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,1536,1.5364835527208116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,1024,1.1965262095133464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,768,0.77474578221639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,768,0.9347662395901151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,1024,1.1398720211452908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,512,0.5826844639248318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,768,0.9236009385850695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,256,0.4865742259555393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,512,0.760407977634006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,512,0.7263742023044162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,128,0.29789688852098256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,256,0.6754622459411621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,256,0.590503109825982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,64,0.2492435508304172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,51200,128,0.417414214875963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,128,0.6341342396206325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,51200,32,0.2594737741682264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,64,0.6528879801432291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,51200,32,0.7479546864827474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,51200,15.205023871527779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,65536,16.654171413845486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,65536,31.280731201171875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,16384,4.639751858181424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,51200,23.76123046875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,65536,47.56749810112847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,16384,7.1016845703125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,16384,11.758352491590712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,12288,3.0368978712293835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,51200,36.97173055013021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,10240,2.650061289469401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,12288,5.358891381157769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,8192,2.470072004530165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,12288,8.846716139051649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,10240,4.461933135986328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,10240,7.381161159939236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,8192,3.588631100124783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,7168,2.12263912624783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,8192,5.81207275390625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,6144,1.6683875189887152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,7168,3.1147261725531687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,7168,4.9796905517578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,5120,1.4092666837904189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,6144,3.06352170308431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,5120,2.0972781711154513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,6144,2.590791914198134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,4096,1.2914631101820204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,5120,2.0452293819851346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,3584,0.9729395972357856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,4096,1.6196791330973308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,4096,1.6201439963446722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,3072,0.7388293478224012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,3584,1.4088933732774522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,3584,1.4161982006496852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,3072,1.1983564164903429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,2560,0.6285973125033909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,3072,1.2338942421807182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,2048,0.5240862104627821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,2560,1.0121706856621635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,2560,1.0422639846801758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,1536,0.4006240102979872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,2048,0.818487114376492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,1024,0.27171021037631565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,2048,0.8585039774576823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,1536,0.6371795336405436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,1536,0.6691146426730685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,768,0.21349777115715873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,1024,0.4686329099867079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,1024,0.4829377598232693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,768,0.35598309834798175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,512,0.15915377934773764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,256,0.11974488364325629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,512,0.29530665609571666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,512,0.2847377724117703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,128,0.10100889205932617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,256,0.21569244066874185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,128,0.1592257817586263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,64,0.08643733130560981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,128,0.2193919950061374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,16384,768,0.38337333997090656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,16384,32,0.08697866731219822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,64,0.22003555297851562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,256,0.24068355560302734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,16384,32,0.24999554951985678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,51200,10.64013671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,65536,12.882089402940538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,65536,26.04891628689236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,16384,3.4643448723687063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,51200,19.525497436523438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,65536,37.23765733506944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,16384,5.918240865071614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,51200,28.92889404296875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,12288,2.4226150512695312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,16384,9.22741953531901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,10240,2.204501258002387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,12288,4.4778594970703125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,10240,3.7388793097602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,10240,5.725443522135417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,8192,1.749217775132921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,12288,6.934895833333333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,7168,1.5198578304714625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,7168,2.4319992065429688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,8192,3.1550532446967234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,8192,3.007542292277018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,6144,1.3547484079996746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,7168,2.6182515886094833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,6144,1.9975208706325953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,6144,2.164055930243598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,5120,1.0874178144666884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,5120,1.6274169286092122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,4096,0.8957227071126302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,4096,1.2938364876641166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,5120,1.733473777770996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,3584,0.878177801767985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,4096,1.3806240293714735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,3072,0.7409831153021919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,3584,1.1356275346544054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,3584,1.2070550918579102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,2560,0.5106319851345485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,3072,0.9746756023830838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,2560,0.8250524202982584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,3072,1.0790328979492188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,2048,0.44977956347995335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,1536,0.3421448866526286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,2560,0.8874959945678711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,2048,0.6717271274990506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,2048,0.7295120027330187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,1024,0.2350346777174208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,1536,0.6091786490546333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,1536,0.5680560006035699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,1024,0.3729333347744412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,1024,0.4036382304297553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,768,0.2855128977033827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,512,0.11719644069671631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,768,0.319814231660631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,512,0.22425511148240831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,512,0.24409688843621147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,256,0.09074133634567261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,768,0.16759022076924643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,256,0.18435288800133598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,128,0.0756124456723531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,256,0.17646755112542045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,64,0.06499022245407104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,12288,128,0.12833155526055232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,128,0.164682666460673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,12288,32,0.06847644514507718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,64,0.16239910655551487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,12288,32,0.17414132754007974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,51200,8.751309712727865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,65536,11.058620029025606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,65536,23.144022623697918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,16384,2.794640858968099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,51200,17.655782063802082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,65536,32.09148491753472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,12288,1.8799137539333766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,16384,5.378378550211589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,16384,7.969963921440972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,51200,24.963485717773438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,10240,1.731841829087999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,12288,4.057857937282987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,10240,3.757409837510851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,12288,5.893007914225261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,10240,3.384768803914388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,8192,1.3891386455959742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,7168,1.3009778128729927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,8192,2.4109280904134116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,8192,2.7269937727186413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,7168,2.054729673597548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,6144,0.9846987194485135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,5120,0.907133314344618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,7168,2.3752159542507596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,6144,1.7341101964314778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,6144,1.974520789252387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,4096,0.6541031201680502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,5120,1.4451298183865016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,5120,1.5915572908189561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,4096,1.1475849151611328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,3584,0.5428053538004557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,4096,1.2680933210584853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,3072,0.5106719864739312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,3584,1.0118320253160265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,3584,1.106646219889323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,3072,0.8726951281229655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,2560,0.42380711767408585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,2048,0.378290679719713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,3072,0.9602835973103842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,2560,0.7362906667921277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,2560,0.8114328914218478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,1536,0.27750221888224286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,2048,0.7150897979736328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,2048,0.6657715373569065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,1536,0.46573601828681094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,1024,0.1893528832329644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,1536,0.5182986789279515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,768,0.14696889453464085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,1024,0.3670328987969293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,768,0.252492454316881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,768,0.2883422109815809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,1024,0.3297146691216363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,256,0.07714666922887166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,512,0.19721421930525038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,512,0.2196497784720527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,128,0.06397422154744466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,256,0.15626488791571722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,256,0.16105333964029947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,128,0.14134400420718723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,10240,128,0.1132871045006646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,64,0.053928888506359525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,32,0.05606044663323296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,64,0.1387031078338623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,10240,512,0.11028266615337795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,10240,32,0.14575466844770643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,51200,6.881536695692275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,65536,8.616710238986546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,65536,17.616811116536457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,16384,2.2900808122422958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,51200,13.827974955240885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,51200,18.72106764051649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,65536,26.06005859375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,16384,5.993208143446181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,12288,1.7148906919691298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,16384,4.439518398708768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,10240,1.6788711547851562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,12288,3.5138265821668835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,10240,2.6441182030571833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,12288,3.3338328467475047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,8192,1.2461609310574002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,7168,1.0793271594577365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,10240,2.7496577368842234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,8192,2.198452419704861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,8192,2.0211618211534286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,7168,1.7707911597357855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,6144,0.703562683529324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,7168,1.9282915327284071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,5120,0.7215110990736219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,6144,1.5074355867173936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,6144,1.6615982055664062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,4096,0.5627982351515028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,5120,1.2639395395914714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,5120,1.391301367017958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,3584,0.5093599955240885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,4096,1.005616929796007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,4096,1.1407048967149522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,3072,0.3885795540279812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,3584,0.8862711058722602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,2560,0.3511875470479329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,3072,0.7653351359897189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,3584,1.0234133402506511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,3072,0.8669048945109049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,2048,0.29333066940307617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,2560,0.6500239902072483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,1536,0.2104426622390747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,2048,0.5274266666836208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,2048,0.6010115411546495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,1536,0.41286667188008624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,1024,0.1477457814746433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,1536,0.46750402450561523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,768,0.12133688396877712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,1024,0.2865911059909397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,1024,0.33109866248236763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,768,0.21812000539567736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,512,0.08526488807466294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,2560,0.73489777247111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,768,0.2581804328494602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,256,0.06271911329693265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,512,0.1695128944185045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,128,0.05274577935536703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,256,0.13493866390652126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,256,0.13686933782365587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,512,0.19874400562710234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,64,0.0455662210782369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,8192,32,0.04893777767817179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,128,0.11770755714840359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,8192,128,0.09964355495240952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,32,0.11766489346822102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,8192,64,0.1148480044470893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,51200,6.026741451687283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,65536,7.367738511827256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,65536,16.722391764322918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,16384,2.0091368357340493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,65536,20.946851942274304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,51200,13.119114345974394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,51200,15.995698716905382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,12288,1.4520177841186523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,16384,5.0482177734375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,16384,4.190613216824002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,12288,2.9788710276285806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,12288,3.162472830878364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,10240,1.2325947019788954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,8192,1.0006533728705513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,10240,2.393009821573893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,8192,1.8631680806477864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,10240,2.606166203816732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,7168,0.9437369240654839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,8192,2.0884710947672525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,6144,0.8095955318874783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,7168,1.6370995839436848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,7168,1.8325191073947484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,5120,0.6085217793782552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,6144,1.3987333509657118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,4096,0.4895608690049913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,6144,1.5778035057915583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,5120,1.1762293709648979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,5120,1.3324391047159831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,3584,0.4397599962022569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,4096,0.9487661785549588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,4096,1.0744586520724826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,3072,0.37749067942301434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,3584,0.8254311349656848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,3584,0.949584854973687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,2560,0.2980106671651204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,3072,0.7128231260511609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,3072,0.825924449496799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,2048,0.26412977112664116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,2560,0.6974088880750867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,2560,0.6026097933451334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,2048,0.4910835689968533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,1536,0.19401689370473227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,2048,0.5713306532965766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,1024,0.1292977730433146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,1536,0.44289599524603945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,1536,0.38153865602281356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,1024,0.26551198959350586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,1024,0.31180355283949107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,768,0.10114222102695042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,512,0.0752355522579617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,768,0.2430622312757704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,512,0.1555511156717936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,256,0.05475111140145195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,512,0.18455910682678223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,768,0.20225955380333793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,128,0.04633333285649618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,256,0.12218488587273492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,256,0.12739022572835287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,64,0.03892888956599765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,7168,128,0.09199021922217475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,128,0.10639466842015584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,7168,32,0.040299554665883384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,64,0.10181955496470134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,7168,32,0.10379733641942342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,65536,6.2397308349609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,51200,4.837159050835504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,65536,15.662529839409721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,65536,17.433764987521702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,16384,1.6863021850585938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,51200,12.251892937554253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,51200,13.482890658908419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,12288,1.2222649256388347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,16384,4.158250596788195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,12288,2.665777842203776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,16384,3.9435929192437063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,10240,1.1691040462917752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,8192,0.8247466617160373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,12288,2.9613804287380643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,10240,2.1765865749782987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,10240,2.453148523966471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,7168,0.7622906896803113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,8192,1.7097608778211806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,8192,1.9716436598036025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,6144,0.6279999944898818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,7168,1.510718239678277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,7168,1.7293795479668512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,6144,1.295474688212077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,5120,0.5482639736599392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,6144,1.4895938237508137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,4096,0.4628035757276747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,5120,1.0836328930324979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,5120,1.254277335272895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,4096,0.8683555391099719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,3584,0.39035820960998535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,4096,1.0264017317030165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,3072,0.35950756072998047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,3584,0.7634231249491373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,3584,0.8946053187052408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,2560,0.2692462338341607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,3072,0.6605520248413086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,3072,0.91940795050727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,2048,0.21962666511535645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,2560,0.5576497713724772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,2560,0.6584026548597548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,1536,0.16474578115675184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,2048,0.4585537910461426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,1024,0.11674222681257461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,2048,0.5386151207817925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,1536,0.3522666560278998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,1536,0.41683112250434023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,768,0.09135111172993977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,1024,0.24386133088005912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,512,0.06782399945788913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,768,0.18530666828155518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,1024,0.29246044158935547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,768,0.22714045312669542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,256,0.048341333866119385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,512,0.14156799846225315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,512,0.17143555482228598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,128,0.04042399923006693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,256,0.10936799976560806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,64,0.03828977876239353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,256,0.11764532989925808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,128,0.09391288624869452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,6144,128,0.08407821920182969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,6144,32,0.04145866632461548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,64,0.08958488702774048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,6144,32,0.09000266922844781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,51200,4.069550408257379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,65536,5.4313769870334205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,65536,14.753078884548612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,65536,14.705291748046875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,51200,11.459588792588976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,16384,1.255375968085395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,51200,11.508895026312935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,12288,1.058177736070421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,16384,3.3996374342176647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,16384,3.6969947814941406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,10240,0.8078373273213705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,12288,2.3986434936523438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,12288,2.766317367553711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,8192,0.6223048633999294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,10240,1.9766559600830078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,8192,1.5763173633151586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,10240,2.3079537285698786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,7168,0.6126586596171061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,8192,1.8533848656548395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,6144,0.5502613385518392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,7168,1.3899058236016169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,6144,1.1922355228000217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,7168,1.6312035454644098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,5120,0.4387164380815294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,4096,0.35601422521803117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,6144,1.4019528494940863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,5120,0.9977893829345703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,5120,1.1807102627224393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,4096,0.8010426627265083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,3584,0.3177119890848796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,4096,0.9551359812418619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,3072,0.24064800474378797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,3584,0.7047368685404459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,3584,0.843016889360216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,3072,0.6344301965501573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,2560,0.23402312066819933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,3072,0.7323795424567329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,2048,0.17954666084713408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,2560,0.5134062237209743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,2560,0.618151134914822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,1536,0.1389662159813775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,2048,0.4178737799326579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,1536,0.3232604397667779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,1024,0.09863555431365967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,1536,0.3912915653652615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,768,0.07639733288023207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,1024,0.22235022650824654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,1024,0.2713724507225884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,768,0.16809244950612387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,512,0.056095110045539014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,768,0.2108222246170044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,256,0.04055644406212701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,512,0.1275093290540907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,512,0.15763378143310547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,128,0.03322666552331712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,256,0.10668177737130059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,2048,0.50600798924764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,128,0.08210577567418416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,64,0.0298942228158315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,5120,128,0.07607466644710965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,5120,32,0.031530665026770696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,64,0.07698666387134127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,32,0.07668533590104845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,5120,256,0.0972471104727851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,51200,3.4182781643337674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,65536,4.435140397813585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,65536,12.665701972113716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,16384,0.9765342076619467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,65536,13.729327731662325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,51200,10.189234415690104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,12288,0.7678177621629504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,51200,10.720710754394531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,16384,2.9786764780680337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,10240,0.6985217730204264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,12288,2.170901404486762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,16384,3.4594400193956165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,12288,2.5883360968695746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,8192,0.5798720253838433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,10240,1.8005672030978734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,8192,1.439009772406684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,10240,2.161921819051107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,7168,0.5201413366529677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,8192,1.739007102118598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,6144,0.4728879928588867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,7168,1.267312897576226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,5120,0.32875998814900714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,7168,1.5289928648206923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,6144,1.0868302451239693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,4096,0.2677146593729655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,6144,1.315599971347385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,5120,0.9122248755560981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,5120,1.1065137651231554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,4096,0.7281004587809244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,3584,0.22618844774034289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,4096,0.8937991460164388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,3072,0.19320621755388048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,3584,0.6415031221177843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,3584,0.7882062064276801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,2560,0.1725848913192749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,3072,0.5547786818610297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,2048,0.14707556035783556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,3072,0.6856906678941516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,2560,0.4682115448845758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,1536,0.11176444424523248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,2560,0.5783217748006185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,2048,0.38131732410854763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,1024,0.07923199733098348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,2048,0.4730524486965603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,1536,0.2940311166975233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,1536,0.36525599161783856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,768,0.06073333157433403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,1024,0.20142577754126656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,512,0.044943110810385815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,1024,0.252799113591512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,768,0.15194843875037298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,768,0.1941359970304701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,256,0.03367288907368978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,256,0.08414222134484185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,256,0.09684889184104072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,128,0.028132445282406274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,512,0.14503378338283962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,64,0.02401155564520094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,4096,128,0.06824622551600139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,128,0.07044088840484619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,4096,32,0.024431111084090337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,32,0.06382133563359578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,64,0.06515110863579644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,4096,512,0.11355555057525635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,65536,4.069051530626085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,51200,2.907774183485243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,65536,11.829026963975695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,16384,0.8378862275017632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,51200,9.569040086534288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,65536,13.28858862982856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,12288,0.7026471032036675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,16384,2.8130416870117188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,51200,10.40590582953559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,16384,3.3474951850043406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,10240,0.57768800523546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,12288,2.052745819091797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,8192,0.45598755942450625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,12288,2.511593712700738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,10240,1.71471553378635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,8192,1.369192017449273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,7168,0.37127288182576496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,10240,2.0993164910210504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,6144,0.3752426571316189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,7168,1.2057261996799045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,8192,1.6882923973931208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,5120,0.3082204394870334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,7168,1.4813466601901586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,6144,1.035896937052409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,4096,0.2202319966422187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,6144,1.276221381293403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,5120,0.868135134379069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,4096,0.6950444645351834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,5120,1.0723351372612846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,3584,0.21120444933573404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,3072,0.1783297724193997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,3584,0.61190398534139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,4096,0.8672800064086914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,3584,0.7643146514892578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,2560,0.1661306619644165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,3072,0.5280728869967991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,3072,0.6643297937181261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,2048,0.1283608939912584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,2560,0.5591413180033366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,1536,0.0896844466527303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,2048,0.36364976565043133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,2048,0.4574115541246202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,1024,0.06663733058505587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,1536,0.2802373303307427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,1536,0.35354312260945636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,768,0.05414577656322055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,1024,0.24194222026401094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,2560,0.44626580344306094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,1024,0.19113066461351183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,512,0.040561778677834406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,768,0.14351377222273085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,256,0.02992977698644002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,768,0.18662310971154106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,512,0.1379635598924425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,128,0.025115556187099878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,256,0.0784097777472602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,512,0.10710666577021281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,256,0.0915457804997762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,64,0.021720889541837905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,128,0.06421422296100192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3584,32,0.02202044427394867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3584,128,0.06481511063045926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,64,0.057908442285325795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3584,32,0.056927111413743764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,51200,2.878485361735026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,65536,3.441504796346029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,65536,11.393555535210503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,51200,8.862179226345486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,16384,0.7989741961161295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,65536,12.882502237955729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,51200,10.079644097222223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,12288,0.6043359968397353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,16384,2.644438213772244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,10240,0.506698661380344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,16384,3.240406248304579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,12288,1.951967027452257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,8192,0.40659912427266437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,12288,2.4320390489366317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,10240,1.6296479966905382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,7168,0.37801244523790145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,8192,1.301356421576606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,10240,2.0358685387505426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,6144,0.3205840057796902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,8192,1.6343350940280492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,7168,1.1484195921156142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,5120,0.2768293221791585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,6144,0.9837812847561307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,7168,1.438637309604221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,4096,0.2021155489815606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,6144,1.2369172837999132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,5120,0.825891547732883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,3584,0.19086221853892008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,5120,1.0403128729926217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,4096,0.6595413419935439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,4096,0.840261353386773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,3072,0.1545742220348782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,3584,0.585736009809706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,2560,0.1390648947821723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,3072,0.5017546547783746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,3584,0.7419857978820801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,3072,0.6428186628553603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,2048,0.11313777499728733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,2560,0.4235564337836371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,2560,0.5424604415893555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,1536,0.0862506694263882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,2048,0.44161778026156956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,1536,0.26582309934828013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,1024,0.061016890737745494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,1536,0.3404248820410834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,1024,0.18029244740804037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,768,0.04505866765975952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,1024,0.23321776919894746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,768,0.1352871126598782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,512,0.036109334892696805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,768,0.17895821730295816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,2048,0.34569509824117023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,256,0.0266275554895401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,512,0.099199997054206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,512,0.1322986682256063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,128,0.02239022155602773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,256,0.0870400004916721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,128,0.058515555328793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,64,0.019271110494931538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,256,0.07208000289069282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,3072,32,0.019863999552196927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,3072,128,0.06050044298171997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,64,0.051923556460274585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,3072,32,0.05080799924002754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,51200,2.173113716973199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,65536,2.81111208597819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,65536,10.585261874728733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,51200,8.151005215115017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,16384,0.648304886288113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,51200,9.672525193956163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,65536,12.356201171875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,12288,0.5106497870551215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,16384,2.4899929894341364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,10240,0.44593334197998047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,16384,3.106626722547743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,12288,1.8450489044189453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,8192,0.33052534527248806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,10240,1.5411742528279622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,12288,2.341109381781684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,7168,0.29146843486362034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,8192,1.2332302729288738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,10240,1.9549475775824652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,6144,0.28041066063774955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,7168,1.0838151507907443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,8192,1.5736515257093642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,5120,0.20146221584743926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,7168,1.3799279530843098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,6144,0.9347155888875326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,4096,0.17854932943979898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,6144,1.1907662285698783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,5120,0.7810666826036242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,3584,0.15153600109948054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,4096,0.6257875760396322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,5120,0.9978230794270834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,3072,0.1446559958987766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,4096,0.8073582119411893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,3584,0.551471127404107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,3584,0.7103146447075738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,3072,0.4762168990241156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,3072,0.6161084704928929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,2560,0.40175289577907985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,2048,0.08932711018456353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,2560,0.51856443617079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,1536,0.06978933016459148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,2048,0.3262088828616672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,2048,0.4241200023227268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,1536,0.2509182294209798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,1024,0.04833955566088358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,2560,0.10778755611843532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,1536,0.32563643985324436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,768,0.039451556073294744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,1024,0.16932266288333467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,1024,0.22217867109510633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,768,0.126129772928026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,512,0.09214399920569526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,256,0.022251556317011516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,512,0.12500177489386663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,256,0.06510844495561388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,256,0.08159555329216851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,128,0.019726221760114033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,128,0.05210666524039375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,128,0.05671022335688273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,64,0.0192248887485928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,32,0.019877332780096266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,64,0.045935998360315956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2560,512,0.029250668154822454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2560,32,0.04351644383536445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2560,768,0.17140800423092314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,51200,1.871154573228624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,65536,2.3794267442491317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,65536,9.535964118109808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,51200,7.5386962890625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,16384,0.5514479743109809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,51200,9.360237969292536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,65536,11.962847391764322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,16384,2.3329075707329645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,12288,0.4340497652689616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,16384,3.0071199205186634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,10240,0.33442489306132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,12288,1.7422960069444444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,8192,0.2719279925028483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,12288,2.264478259616428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,10240,1.4516284730699327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,7168,0.2327368789248996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,8192,1.1639475292629666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,10240,1.8949510786268446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,6144,0.2015742195977105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,7168,1.0233484903971355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,8192,1.5210364659627278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,5120,0.17932355403900146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,7168,1.3375155131022136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,6144,0.880673779381646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,5120,0.735231982337104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,4096,0.13762488630082872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,6144,1.1501128938462999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,3584,0.11969244480133057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,4096,0.5907528665330675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,5120,0.967042711046007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,3584,0.5199315812852647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,4096,0.7811466852823893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,3072,0.102783997853597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,2560,0.08500888612535264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,3072,0.4496480094061957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,3584,0.6863386895921496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,2048,0.07967555522918701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,2560,0.37995288107130265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,3072,0.5953066613939073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,1536,0.05836888816621569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,2048,0.30797065628899467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,2560,0.5007964240180122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,1024,0.040528890159395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,2048,0.4085182348887126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,1536,0.2362204392751058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,1536,0.3140559991200765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,1024,0.15792977809906006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,768,0.032568001084857516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,1024,0.21311644713083902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,512,0.023638221952650283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,768,0.11782222323947483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,768,0.16242489549848768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,512,0.08504266871346368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,256,0.017844445175594754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,128,0.014800000521871777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,512,0.11926755640241836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,256,0.059134225050608315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,256,0.07708711094326444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,64,0.01441511180665758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,2048,32,0.015221332510312399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,64,0.039232889811197914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,2048,128,0.05290666553709242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,32,0.0370124446021186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,2048,128,0.04587733414438036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,65536,1.756717363993327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,51200,1.3989431593153212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,51200,6.929214901394314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,65536,8.858105977376303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,16384,0.4228737884097629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,16384,2.1811190711127386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,51200,9.04917229546441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,12288,0.3332773314581977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,65536,11.56782446967231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,10240,0.27284267213609487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,16384,2.9135903252495656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,12288,1.6338266796535914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,10240,1.363968849182129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,8192,0.2277031209733751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,12288,2.194003634982639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,10240,1.8322861989339192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,7168,0.18694667021433511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,8192,1.0941990746392143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,6144,0.1511733399497138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,7168,0.9607466591729058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,8192,1.4708870781792536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,5120,0.13338400257958308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,7168,1.2931982676188152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,6144,0.8263031111823188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,4096,0.10124711195627849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,5120,0.6907173262702094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,6144,1.113343980577257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,3584,0.09308622280756633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,4096,0.5559155676099989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,5120,0.9342560238308377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,3072,0.07715377542707655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,4096,0.7529439926147461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,3584,0.4894311163160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,2560,0.07086133294635348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,3584,0.6621173222859701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,3072,0.42313422097100156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,3072,0.5727279980977377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,2560,0.3563431104024251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,2048,0.058393776416778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,1536,0.04445777667893303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,2560,0.4830800162421332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,2048,0.2881128787994385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,2048,0.3936533398098416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,1024,0.03162577748298645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,1536,0.3018728892008464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,1536,0.2205359935760498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,768,0.024655999408827886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,1024,0.14595911237928602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,1024,0.20341689056820342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,512,0.018383999665578205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,768,0.15448089440663657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,768,0.10843733284208511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,256,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,512,0.07837066385481092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,512,0.11221778392791748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,128,0.01237155579858356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,256,0.07198577457004122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,256,0.05232533481385973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,128,0.0384844442208608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,64,0.01147822207874722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1536,128,0.04788711004787021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1536,32,0.013463111387358772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,64,0.03193688723776076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1536,32,0.029860443539089624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,65536,1.1480133268568251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,51200,0.8754017617967395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,51200,6.388251410590278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,16384,0.28001422352261013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,65536,8.150419447157118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,16384,2.0370542738172746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,12288,0.21287200185987684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,51200,8.648993598090277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,16384,2.7863936954074435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,65536,11.058286878797743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,10240,0.18584622277153862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,12288,1.5312009387546117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,8192,0.141903109020657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,10240,1.2758418189154732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,12288,2.0951830546061196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,7168,0.12958399454752603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,10240,1.7521573172675238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,8192,1.025503158569336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,6144,0.10977600018183391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,7168,0.8988559510972766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,8192,1.4051013522677953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,5120,0.09292533662584092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,6144,0.773753801981608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,7168,1.2361662122938368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,4096,0.07524710893630981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,5120,0.6465697818332249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,6144,1.0620311101277669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,3584,0.06544088655047946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,4096,0.5212888717651367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,5120,0.8917377259996203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,3072,0.057448003027174205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,3584,0.4573342005411784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,4096,0.7185493575202094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,3584,0.6328497992621528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,2560,0.0506586664252811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,3072,0.39516711235046387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,2048,0.042430222034454346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,3072,0.5462817615932888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,2560,0.33231910069783527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,1536,0.03036710951063368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,2560,0.45995553334554035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,2048,0.37375556098090273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,1536,0.20510933134290907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,2048,0.2699262301127116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,1536,0.2866702344682482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,768,0.016494222813182406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,1024,0.13453777631123862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,768,0.09918222162458633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,1024,0.1920808951059977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,512,0.012383999923865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,768,0.14549599753485784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,256,0.010747555229398938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,512,0.10439911153581406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,512,0.06948088937335543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,256,0.04425155454211765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,256,0.06707377566231622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,128,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,64,0.009024889104896123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,128,0.03161333335770501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,32,0.010013333625263637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,1024,128,0.04316622349951002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,1024,1024,0.021307556165589228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,32,0.023990222149425085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,1024,64,0.026147555973794725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,65536,0.9190906948513455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,51200,0.6705804400973849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,51200,6.15060551961263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,16384,0.21586310863494873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,65536,7.856446160210504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,16384,1.9676827324761286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,51200,8.557950337727865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,12288,0.16518489519755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,16384,2.751058578491211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,10240,0.14512622356414795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,65536,10.940038045247396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,12288,1.4802142249213324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,8192,0.12079021665785049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,12288,2.070152918497721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,10240,1.2340870963202584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,8192,0.9910248650444878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,7168,0.10564800103505452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,10240,1.7303733825683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,8192,1.3885218302408855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,7168,0.869816886054145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,6144,0.09048089053895737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,7168,1.2200489044189453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,5120,0.07299821906619601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,6144,0.7480426894293891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,4096,0.057632002565595836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,6144,1.049324459499783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,5120,0.6254720158047146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,3584,0.053103112512164645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,4096,0.5040373272365993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,5120,0.8787991205851237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,3072,0.04689777890841166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,4096,0.7089946534898547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,3584,0.4427519904242621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,2560,0.03888444436921014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,3072,0.38187111748589414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,3584,0.624492433336046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,2560,0.3211839993794759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,3072,0.5389404296875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,2048,0.25987288686964244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,2560,0.4541021982828776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,1536,0.024682665864626568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,2048,0.3684871196746826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,1024,0.017998221847746108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,1536,0.19781955083211264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,1536,0.28163377443949383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,1024,0.12999378310309515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,768,0.014387556248241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,1024,0.18778488371107313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,768,0.09464444716771443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,512,0.010949333508809408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,2048,0.03239999877081977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,256,0.011011555790901184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,512,0.06618311007817586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,512,0.10195466544893052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,128,0.009680888719028896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,256,0.06409421894285414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,256,0.04010222355524699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,128,0.028543998797734577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,64,0.009347555538018545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,128,0.0410773323641883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,768,32,0.009720000127951304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,64,0.023079999619060095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,768,32,0.02065333392884996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,768,768,0.14272267288631862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,65536,0.6093804571363661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,51200,0.46871111128065324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,51200,5.931128184000651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,16384,0.14354933632744685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,65536,7.5730743408203125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,16384,1.9015733930799696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,12288,0.1105031106207106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,51200,8.354933844672308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,16384,2.686205334133572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,12288,1.4291262096828883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,10240,0.10285866922802395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,65536,10.677980211046007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,8192,0.0899697807100084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,12288,2.0232925415039062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,10240,1.193062252468533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,8192,0.9563830693562826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,7168,0.0741457806693183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,10240,1.6879022386338975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,6144,0.06344533628887601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,7168,0.8387751049465604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,8192,1.3571013344658747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,5120,0.05462755428420173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,6144,0.7214853498670789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,7168,1.1918630599975586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,4096,0.04530844423505995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,5120,0.6037297778659397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,6144,1.0242399639553494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,3584,0.040891554620530866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,4096,0.48644087049696183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,5120,0.8584515783521863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,3072,0.03479199939303928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,4096,0.6932133568657769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,3584,0.4266791078779433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,2560,0.028902222712834675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,3072,0.36851021978590226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,3584,0.6098960240681967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,2048,0.02327555583582984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,2560,0.30960355864630806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,3072,0.5266790919833714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,1536,0.018357333209779527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,2560,0.4429724481370714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,2048,0.2514977720048693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,2048,0.35894044240315753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,1536,0.19082755512661406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,1024,0.013060444758998023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,1536,0.27430134349399143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,768,0.010766222245163388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,1024,0.12450133429633246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,1024,0.18268356058332655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,512,0.008736888567606607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,768,0.0905875563621521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,768,0.1381937795215183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,256,0.0069422221018208405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,512,0.06266133652793036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,512,0.09841599729326035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,128,0.006268444574541516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,256,0.05924444728427463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,256,0.03573244478967454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,128,0.025851555996470984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,64,0.005674666828579373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,512,128,0.03939022289382087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,512,32,0.00811111099190182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,64,0.020326221982638042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,512,32,0.018002667360835604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,65536,0.360642671585083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,51200,0.2817395528157552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,51200,5.715498606363933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,16384,0.10278577937020196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,65536,7.300671047634548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,16384,1.8335484398735895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,12288,0.07363110780715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,51200,8.149830288357204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,16384,2.6215582953559027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,10240,0.06297422117657132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,12288,1.3787173165215387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,65536,10.429410298665365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,8192,0.0527724458111657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,10240,1.1506471633911133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,8192,0.9232159720526801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,7168,0.045132444964514844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,10240,1.6498906877305772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,12288,1.9732382032606337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,6144,0.03997866643799676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,7168,0.8095964325798882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,8192,1.32242308722602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,6144,0.6954595777723523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,5120,0.03478577733039856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,7168,1.1628506978352864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,4096,0.028860443168216284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,5120,0.5824444558885362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,6144,1.0002835591634114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,3584,0.025729777084456548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,5120,0.8372720082600912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,4096,0.4690702226426866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,3072,0.02148088812828064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,4096,0.6753342416551379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,3584,0.4118764400482178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,2560,0.01556799974706438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,3584,0.5941360261705186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,3072,0.35559643639458555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,2048,0.013119999733236102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,3072,0.5123270882500542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,2560,0.4312444527943929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,1536,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,2048,0.24198932117886016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,2048,0.3497466776106093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,1024,0.008338666624493068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,1536,0.1835146612591214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,1024,0.11894488334655762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,1536,0.2671395672692193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,768,0.007350222104125553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,2560,0.2983066770765516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,768,0.13307555516560873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,768,0.08706844515270656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,512,0.05953511264589098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,256,0.0053031109273433685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,512,0.09487289190292358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,256,0.03240266773435805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,128,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,256,0.05695111221737332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,128,0.022525333695941504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,64,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,128,0.0370488895310296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,256,1024,0.17533600330352783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,32,0.005299555758635203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,64,0.017224889662530687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,256,32,0.016866667403115165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,256,512,0.006262222097979651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,65536,0.32613955603705513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,51200,0.25331465403238934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,51200,5.622198316786025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,16384,0.09387733538945515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,65536,7.1899507310655375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,16384,1.807568868001302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,12288,0.06616889105902778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,51200,8.087464226616754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,12288,1.3600337770250108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,10240,0.05624710851245456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,16384,2.603507571750217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,65536,10.337782118055555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,8192,0.046216888560189136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,12288,1.9605511559380426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,10240,1.1364427142673068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,7168,0.04042488998836941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,8192,0.9107537799411349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,10240,1.635093371073405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,6144,0.03547822104560004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,8192,1.3144186867607963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,7168,0.7996142175462512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,5120,0.030272000365787085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,6144,0.686997307671441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,7168,1.1514772839016385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,6144,0.9906764560275607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,4096,0.02476177778508928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,5120,0.5749715699089898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,4096,0.462465763092041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,3584,0.022601778308550518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,4096,0.6678880055745443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,3584,0.4063013394673665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,3072,0.014284445179833306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,3584,0.5860533184475368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,3072,0.35073065757751465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,2560,0.009658666948477427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,5120,0.8289830949571398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,3072,0.5054008695814345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,2560,0.2939235634273953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,2048,0.0083155557513237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,2560,0.42455289098951554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,1536,0.007262222468852997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,2048,0.2376177840762668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,2048,0.34417155053880477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,1024,0.005921777751710679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,1536,0.1799617740843031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,768,0.005219555563396878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,1536,0.26270755132039386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,1024,0.11634222666422527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,512,0.004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,768,0.0847511092821757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,768,0.1309057739045885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,256,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,512,0.05780000156826443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,512,0.09281777673297459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,128,0.003896000070704354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,1024,0.17280266020033094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,256,0.031000001562966242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,256,0.054823110500971474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,64,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,128,32,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,128,0.02102222210831112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,64,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,128,32,0.014484445254007975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,65536,0.32112712330288357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,51200,0.2493306795756022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,16384,0.09126844671037461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,16384,1.7971227433946397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,12288,0.06469066937764485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,12288,1.3512008455064561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,10240,0.05476088987456428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,10240,1.1287600199381511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,8192,0.04460710949367947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,51200,5.583828396267361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,7168,0.03910222318437364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,8192,0.9054142634073893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,6144,0.0347013341055976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,7168,0.7938471370273166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,5120,0.030322667625215318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,6144,0.6826435724894205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,4096,0.024128889044125874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,65536,7.149304707845052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,3584,0.022514666120211285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,5120,0.5708835389879015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,4096,0.45920711093478733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,3072,0.012736000120639801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,2560,0.008979555633332994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,3584,0.40331021944681805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,2048,0.008994667066468133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,3072,0.347961770163642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,2560,0.29144710964626735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,1536,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,2048,0.234827544954088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,1024,0.006329777754015393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,768,0.005428444594144821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,1536,0.17822666962941489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,1024,0.11530933115217422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,768,0.08349066972732544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,512,0.004892444445027245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,256,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,128,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,512,0.056511110729641385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,256,0.030272887812720403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,64,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,64,32,0.0035128887328836652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,128,0.020280889338917203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,64,0.015453333655993143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,64,32,0.0141004439857271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,65536,0.33493688371446395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,51200,0.24911022186279297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,16384,0.08168888754314847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,16384,1.7950221167670355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,12288,0.06184533569547865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,12288,1.3501208623250325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,10240,0.051479998562071055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,10240,1.128551059299045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,8192,0.04235466652446323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,51200,5.586276584201389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,7168,0.037645333343082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,8192,0.9050452974107531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,6144,0.03306577934159173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,7168,0.7944169044494629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,5120,0.0284551117155287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,65536,7.144874572753906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,6144,0.6822960111829969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,4096,0.02302933401531643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,5120,0.5709946420457628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,4096,0.4592915640936957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,3584,0.020598222812016804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,3072,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,3584,0.4068471060858832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,3072,0.3501733409033881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,2560,0.007694222033023834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,2048,0.006950221955776215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,2560,0.2932444413503011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,2048,0.2372586727142334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,1536,0.005923555543025334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,1024,0.004931555440028508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,1536,0.17968977822197807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,1024,0.11701956060197617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,768,0.00453599997692638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,512,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,768,0.08452355861663818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,512,0.05685244666205513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,256,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,128,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,64,0.00314666661951277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,256,0.0302782224284278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16384,32,32,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,64,0.015630222029156156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,128,0.02039377722475264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16384,32,32,0.01346933344999949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16384,128,128,0.0363395545217726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,51200,25.616992526584202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,16384,8.34100087483724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,51200,33.55827840169271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,16384,10.19559054904514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,16384,17.88728502061632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,12288,6.117091708713108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,12288,9.542816162109375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,10240,5.280736711290148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,12288,7.381731669108073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,10240,6.919104682074653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,8192,4.224269443088107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,10240,5.966651492648655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,8192,5.187035454644097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,7168,3.6463699340820312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,8192,4.5556721157497835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,51200,64.10659450954861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,6144,3.111869388156467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,7168,4.488711039225261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,7168,3.9213617112901478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,5120,2.5894879235161676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,6144,3.740637461344401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,6144,3.335840861002604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,5120,3.136370764838325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,4096,2.0375484890407987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,5120,2.808184093899197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,4096,2.4875208536783853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,4096,2.2784623040093317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,3584,1.8440159691704645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,3072,1.6456791559855144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,3584,2.183932410346137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,3072,1.8803075154622395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,3584,1.9972017076280382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,2560,1.323529773288303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,3072,1.7328213585747614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,2560,1.5945990880330403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,2560,1.4636578030056422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,2048,1.2220079633924696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,2048,1.2914080089992948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,1536,0.8465813530815972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,2048,1.2050133811102974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,1536,1.0147031148274739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,1024,0.6332915623982748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,1024,0.7552186648050944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,1536,0.9471395280626086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,768,0.5106933381822374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,1024,0.7116373380025228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,512,0.38708000712924534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,768,0.5893413225809733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,768,0.5968524614969889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,256,0.2397822274102105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,512,0.534621344672309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,128,0.1987253295050727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,512,0.5374115308125814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,256,0.4700977537367079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,256,0.3704933325449626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,64,0.16685867309570312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,65536,128,0.281184885236952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,65536,32,0.18159111340840658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,128,0.4434559875064426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,64,0.46927822960747617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,65536,32,0.4730861981709798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,51200,20.021472507052952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,65536,25.460979885525173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,65536,33.96176486545139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,16384,6.403939565022786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,51200,26.78183152940538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,16384,14.15338389078776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,12288,4.738202836778429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,16384,8.15466054280599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,65536,66.00158013237846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,51200,49.803253173828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,12288,7.647153218587239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,12288,5.977723439534505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,10240,4.858473883734809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,10240,5.540673573811849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,8192,3.270609749688042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,10240,4.7872772216796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,8192,4.159559885660808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,7168,3.3925342559814453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,8192,3.691588507758247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,7168,3.610334184434679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,6144,2.422337849934896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,7168,3.1852631039089627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,6144,2.9935663011338978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,5120,2.3245555029975042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,6144,2.6780959235297312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,5120,2.5174872080485025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,5120,2.266662173800998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,4096,1.8409894307454426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,4096,1.9920525021023219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,3584,1.6213262346055772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,4096,1.8242710961235895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,3584,1.7478435304429796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,3072,1.267716407775879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,3584,1.6030071046617296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,3072,1.508435567220052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,3072,1.4058684243096247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,2560,1.1200844446818035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,2560,1.2792764239841037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,2560,1.1938560273912218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,2048,0.8950444327460395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,2048,1.0333324008517795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,1536,0.653240892622206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,2048,0.9793128967285156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,1536,0.9918008380466037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,1024,0.4422960016462538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,1536,0.766849782731798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,768,0.34713511996799046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,1024,0.6050711207919651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,1024,0.5761849085489908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,768,0.4701724582248264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,512,0.2603360017140706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,256,0.1876880062950982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,768,0.4768666691250271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,512,0.3965173297458225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,512,0.37436887953016496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,128,0.1531511147816976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,256,0.3677546713087294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,128,0.21942310863071016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,64,0.14007555113898382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,128,0.33035466406080455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,51200,256,0.28906933466593426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,51200,32,0.1443813376956516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,64,0.34060711330837673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,51200,32,0.35404443740844727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,51200,6.557215796576606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,65536,8.526664733886719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,65536,13.63932630750868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,16384,2.509856965806749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,51200,10.454891628689236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,16384,5.128208160400391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,65536,23.458338419596355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,51200,17.127975463867188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,16384,3.289308547973633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,12288,1.802536858452691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,10240,1.5332169002956813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,12288,2.9093360900878906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,12288,2.437316470675998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,10240,2.143658743964301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,8192,1.068518214755588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,10240,2.02274407280816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,7168,1.0017439524332683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,8192,1.6294275919596355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,8192,1.6032080120510523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,6144,0.8038284513685438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,7168,1.4188604354858398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,7168,1.3835182189941406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,5120,0.628661314646403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,6144,1.191989368862576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,6144,1.1823253631591797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,4096,0.46759912702772355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,5120,1.001365343729655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,5120,1.0022044711642795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,3584,0.45150311787923175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,4096,0.797269344329834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,4096,0.8218746715121799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,3584,0.6984711223178439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,3584,0.7181199921502007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,3072,0.6023991372850206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,2560,0.3279866642422146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,3072,0.6280888981289333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,2560,0.5094898011949327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,2048,0.266795555750529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,3072,0.37665332688225645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,2560,0.5315333472357856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,2048,0.4097564485337999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,1536,0.21508177121480307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,1536,0.31648622618781197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,1024,0.153912001185947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,1536,0.33480000495910645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,1024,0.23309066560533312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,1024,0.24709865781995985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,768,0.11492178175184463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,768,0.2023226684994168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,768,0.18267377217610678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,512,0.08149333132637872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,2048,0.4302862220340305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,256,0.061956445376078285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,512,0.15097244580586752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,256,0.11884178055657281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,256,0.12958489523993597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,128,0.05212355653444926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,128,0.11728799343109131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,128,0.090737779935201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,64,0.04601688848601448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,16384,32,0.048841777775022716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,16384,512,0.1591386662589179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,64,0.11935110886891682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,16384,32,0.12318310472700332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,51200,5.107589297824435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,65536,6.077738867865668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,65536,11.193761189778646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,16384,1.6992435455322266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,51200,8.762285868326822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,65536,18.553504096137154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,16384,4.074439154730903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,51200,13.475352817111544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,16384,2.740944968329536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,12288,1.2339395946926541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,10240,1.0760399500528972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,12288,2.3396515316433377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,12288,2.0773332383897567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,10240,1.7503119574652777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,8192,0.8513555526733398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,10240,1.7283315658569336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,8192,1.3356889088948567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,7168,0.748386647966173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,8192,1.3676284154256184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,7168,1.1637360254923503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,6144,0.6174035602145725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,5120,0.5566293398539225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,7168,1.1859448750813801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,6144,0.9812471601698133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,6144,1.0182231267293294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,4096,0.380221340391371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,5120,0.855955547756619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,5120,0.8253333303663465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,4096,0.6558897760179307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,3584,0.3513004514906142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,4096,0.6941884358723959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,3072,0.30299822489420575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,3584,0.5762044588724772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,3584,0.6166675355699327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,2560,0.257712894015842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,3072,0.496687995062934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,3072,0.5368417633904351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,2048,0.22287199232313368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,2560,0.4200266732109918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,1536,0.1614453395207723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,2560,0.45678043365478516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,2048,0.3367662164900038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,2048,0.3807635572221544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,1024,0.11430222458309597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,1536,0.2727102173699273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,768,0.08527644475301106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,1536,0.2860177887810601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,1024,0.20916088422139487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,1024,0.18957155280643037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,512,0.06268977456622653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,768,0.1529031064775255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,256,0.048356443643569946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,768,0.17115644613901773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,512,0.13392533196343318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,512,0.12221954928504096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,128,0.04051555527581109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,256,0.1024177802933587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,64,0.038378665844599404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,256,0.09948888752195571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,12288,128,0.0751075545946757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,128,0.0922657781177097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,12288,32,0.04150844282574124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,64,0.09314577447043525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,12288,32,0.09639022085401748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,51200,4.230377621120877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,65536,5.029246012369792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,65536,10.126468234592014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,16384,1.4022720124986436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,51200,7.825374179416233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,51200,11.57081264919705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,65536,15.980941772460938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,16384,3.542873806423611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,16384,2.4946471320258246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,12288,1.0716782675849068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,10240,0.8998532825046115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,12288,2.066227595011393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,12288,1.8844596015082464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,10240,1.5478915108574762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,8192,0.6621208720737034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,10240,1.569792005750868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,7168,0.6076044506496853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,8192,1.1860213809543185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,8192,1.250314712524414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,7168,1.0350275039672852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,6144,0.519428465101454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,7168,1.0822675493028429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,5120,0.41018043624030215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,6144,0.8755093150668674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,6144,0.9360195795694987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,4096,0.331608878241645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,5120,0.7623359892103406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,5120,0.7820124626159668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,3584,0.3125946786668566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,4096,0.5936355590820312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,4096,0.6368862258063422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,3072,0.23647022247314453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,3584,0.5125670962863499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,3072,0.44370222091674805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,3584,0.562020460764567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,2560,0.20044355922275117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,2048,0.15892799695332846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,2560,0.3750444518195258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,2560,0.414955563015408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,3072,0.49088621139526367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,1536,0.13065332836574978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,2048,0.29977332221137154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,1024,0.09970577557881673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,2048,0.33481600549485946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,1536,0.25846044222513836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,768,0.06775466601053874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,1024,0.16799022091759575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,1024,0.18994577725728354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,512,0.05575999948713514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,1536,0.23024890157911512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,768,0.13151289357079401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,768,0.1544817818535699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,256,0.04061155517896017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,512,0.10727200243208145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,128,0.03344977895418803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,256,0.08836622370613946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,256,0.0893546673986647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,512,0.12212799655066596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,64,0.029866665601730347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,128,0.07959289020962186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,10240,32,0.03164711263444688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,10240,128,0.06712977753745185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,32,0.0825022194120619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,10240,64,0.08013866345087688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,51200,3.4787688785129123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,65536,4.010313669840495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,65536,9.12043931749132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,16384,1.1573830710517037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,51200,6.950327131483291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,65536,13.475770738389757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,51200,9.624895731608072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,16384,2.2465830908881292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,12288,0.8048826853434244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,16384,3.0127741495768228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,10240,0.7056889004177518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,12288,1.7881991068522136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,10240,1.3544782002766926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,8192,0.5257724655999078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,12288,1.6928337944878473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,7168,0.4786497751871745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,10240,1.4223857455783422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,8192,1.0389671325683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,8192,1.1228649351331923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,6144,0.38005688455369735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,7168,0.9147848553127713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,7168,0.9799608654446073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,5120,0.32587022251553005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,6144,0.7707911067538791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,6144,0.8449733522203234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,4096,0.28076354662577313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,5120,0.6435128847757975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,3584,0.2251866658528646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,5120,0.7254906760321723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,4096,0.5146631134880914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,3072,0.20249511135949028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,4096,0.6035875744289821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,3584,0.45583465364244247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,3584,0.5090079837375218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,2560,0.17689422766367593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,3072,0.3906959957546658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,3072,0.4443511168162028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,2048,0.13812266455756292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,2560,0.32943110995822483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,2560,0.3751342296600342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,1536,0.10342844327290852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,2048,0.26365688112046987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,2048,0.30175288518269855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,1536,0.20009777281019422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,1024,0.07239555650287204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,1536,0.2318728764851888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,768,0.06267466810014513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,1024,0.1460231145222982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,1024,0.1694906685087416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,512,0.04367288947105408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,768,0.1148337788052029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,256,0.03323466579119364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,768,0.13894577821095785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,512,0.09267555342780219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,512,0.10840355687671238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,128,0.02878755662176344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,256,0.07611288626988728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,256,0.0787715580728319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,64,0.02411199940575494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,128,0.06694844696256849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,8192,128,0.06009511152903239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,8192,32,0.025082666012975905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,64,0.0667475528187222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,8192,32,0.06868444548712836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,65536,3.388620376586914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,51200,2.734505759345161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,65536,8.44399939643012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,65536,10.330332438151041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,16384,0.9781164593166776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,51200,7.96486579047309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,12288,0.7368755340576172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,16384,2.1331954532199435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,16384,2.5355493757459855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,51200,6.751185946994358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,10240,0.5547235806783041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,12288,1.5029288397894964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,12288,1.6098569234212239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,10240,1.2035421795315213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,8192,0.4808310932583279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,10240,1.3481591542561848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,7168,0.43607378005981445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,8192,0.9390693240695529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,8192,1.0732737647162545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,7168,0.8269093301561145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,6144,0.3589342170291477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,5120,0.31396177079942494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,7168,0.9316622416178385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,6144,0.7093075646294488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,6144,0.802354653676351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,4096,0.25274401240878636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,5120,0.5986222161187066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,5120,0.6761013136969672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,3584,0.21471733517116973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,4096,0.47656530804104275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,3072,0.1617057720820109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,3584,0.4213537904951308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,3584,0.48447465896606445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,2560,0.16044533252716064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,3072,0.3631262249416775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,3072,0.422452449798584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,2048,0.12458222442203098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,2560,0.30595289336310494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,2560,0.3570151064130995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,1536,0.09849333100848728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,4096,0.6005839771694607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,2048,0.2859102090199788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,2048,0.24553155899047852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,1024,0.06871822145250109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,1536,0.18627199861738417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,768,0.051961776283052236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,1536,0.22045332855648467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,1024,0.16028889020284018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,1024,0.13421421580844456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,768,0.10643021927939521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,512,0.039000001218583845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,768,0.13095911343892416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,512,0.10204088687896729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,512,0.08342044221030341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,256,0.06608888838026258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,128,0.025586666332350835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,256,0.07377155621846516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,64,0.02134311033619775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,128,0.05801244576772054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,32,0.022005332840813532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,7168,128,0.05638933181762695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,64,0.055712891949547656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,7168,32,0.05737777551015218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,7168,256,0.029226667351192896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,51200,2.3585946824815536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,65536,3.2289661831325955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,65536,8.750538296169704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,65536,8.15409681532118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,51200,6.735705905490452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,16384,0.8622968991597494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,51200,6.336307525634766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,12288,0.6935235659281412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,16384,2.092487123277452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,12288,1.3410701751708984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,10240,0.5272657606336806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,16384,2.0251714918348527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,12288,1.5328470865885417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,8192,0.4307013352711995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,10240,1.0938675138685439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,7168,0.37335822317335343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,8192,0.8638773494296603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,10240,1.280151155259874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,8192,1.0174836052788627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,7168,0.7639875411987305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,7168,0.8885919782850477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,6144,0.6572960217793783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,5120,0.2571484512752957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,6144,0.3241742187076145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,6144,0.7653075324164497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,5120,0.5514319737752279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,4096,0.21362222565544975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,5120,0.6415955755445693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,3584,0.18401600254906547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,4096,0.4418799877166748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,4096,0.5227964189317491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,3072,0.1624657842848036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,3584,0.3890311188167996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,3584,0.46087646484375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,2560,0.13151555591159397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,3072,0.3370062245262994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,3072,0.4010053210788303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,2048,0.1094026698006524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,2560,0.28399909867180717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,2560,0.33962755733066136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,1536,0.08514755302005345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,2048,0.2259626653459337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,2048,0.271342224544949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,1024,0.06011466847525703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,1536,0.17180355389912924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,1536,0.20831467045678032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,768,0.046893331739637584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,1024,0.1230435503853692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,1024,0.15164000458187527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,768,0.09784177939097087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,512,0.03423733181423611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,768,0.12340622478061253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,256,0.031256887647840716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,512,0.07589955462349786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,512,0.0965333316061232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,128,0.022449778185950384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,256,0.059528887271881104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,256,0.06872089041603936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,64,0.019277332557572257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,128,0.052060445149739586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,6144,32,0.019835554891162448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,6144,128,0.05247911148601108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,64,0.050069332122802734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,6144,32,0.05063022176424662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,65536,2.6099974314371743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,51200,2.304704030354818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,65536,7.522854275173611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,65536,7.573326958550347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,51200,5.754243638780381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,16384,0.7610177993774414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,12288,0.5691066847907172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,16384,1.7309901979234483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,12288,1.2073964012993705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,16384,1.8852791256374781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,10240,0.42795467376708984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,51200,5.890890757242839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,8192,0.333260456720988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,10240,0.9968942006429037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,12288,1.4247751235961914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,7168,0.3039688799116346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,10240,1.1938355763753254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,8192,0.7974239985148112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,8192,0.948846181233724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,6144,0.28501778178744847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,7168,0.7233431074354383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,7168,0.8303395377265083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,5120,0.23016177283393013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,6144,0.6056097878350152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,4096,0.17117777135637072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,6144,0.7140951156616211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,5120,0.5073351330227321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,3584,0.1484337780210707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,5120,0.6026622454325358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,4096,0.4081839985317654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,4096,0.4892631106906467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,3072,0.14009333981408015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,3584,0.3591769006517198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,3584,0.4322524335649278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,2560,0.10818577475017971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,3072,0.3120746612548828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,3072,0.3781439993116591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,2560,0.26151466369628906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,2048,0.0917102230919732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,2560,0.31729422675238717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,1536,0.07058755556742351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,2048,0.2084275616539849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,1536,0.1569031079610189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,1024,0.046317332320743136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,1536,0.194487108124627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,768,0.03761955433421665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,1024,0.11179378297593857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,1024,0.14040799935658774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,512,0.028001778655582007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,768,0.11427556143866645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,512,0.06911822160085042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,768,0.09056622452206081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,256,0.021839110387696162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,512,0.08919999996821086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,128,0.019918221566412184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,256,0.0635111133257548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,256,0.05351022217008802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,64,0.019241778386963736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,128,0.04582666688495212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,2048,0.2542133331298828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,5120,32,0.019653333557976615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,5120,128,0.048688001102871366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,64,0.044200887282689415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,5120,32,0.043599999613232084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,65536,2.1621439192030167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,51200,1.748445298936632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,65536,6.340593973795573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,16384,0.5980728997124566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,65536,6.925507439507379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,51200,5.11319817437066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,12288,0.471235540178087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,16384,1.483115514119466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,16384,1.7528746922810872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,51200,5.412255181206597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,12288,1.0900755988226998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,10240,0.35782843165927464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,8192,0.27022311422559947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,12288,1.3177999920315213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,10240,0.907216919793023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,10240,1.1018657684326172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,7168,0.2588160037994385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,8192,0.7632257673475477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,6144,0.19865777757432726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,8192,0.8897742165459527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,7168,0.6408471001519097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,7168,0.7808266745673286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,5120,0.1622577773200141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,6144,0.5525849130418565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,4096,0.14025066958533391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,6144,0.6766817834642199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,5120,0.4636524518330892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,5120,0.5694284439086914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,4096,0.37243021859063047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,3584,0.11490844355689155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,3072,0.10277955399619208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,4096,0.46314223607381183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,3584,0.32829154862297905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,3584,0.40954844156901044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,2560,0.09531199932098389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,3072,0.2842328813340929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,3072,0.3557937675052219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,2048,0.0731733308898078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,2560,0.23870934380425346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,2560,0.29955821567111546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,1536,0.058428446451822914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,2048,0.23885154724121094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,2048,0.18960444132486978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,1024,0.04050844576623704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,1536,0.1420960028966268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,768,0.032624890406926475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,1536,0.1829271051618788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,1024,0.1013724472787645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,1024,0.13201599650912815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,512,0.023494222097926672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,768,0.08127378092871772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,768,0.10770222213533188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,256,0.018319111731317308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,512,0.06198133362664116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,128,0.014729777971903482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,512,0.08421955506006877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,256,0.04675199919276767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,256,0.05907999806933933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,64,0.015174221661355762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,128,0.03964089022742377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,4096,32,0.016155555844306946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,4096,128,0.044975999328825206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,64,0.03605955508020189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,4096,32,0.035678221119774714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,51200,1.3755351172553167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,65536,1.717134263780382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,65536,5.929194556342231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,51200,4.6669129265679254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,16384,0.4749457571241591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,65536,6.663155449761285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,51200,5.211155361599393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,12288,0.3560489018758138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,16384,1.4005378087361653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,10240,0.29717334111531574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,12288,1.0360079871283638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,16384,1.6826897727118597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,8192,0.2462915579477946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,12288,1.267006238301595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,10240,0.8631724251641167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,7168,0.20533600118425158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,10240,1.0623618231879342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,8192,0.7436061965094672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,6144,0.18305688434176973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,8192,0.856695122188992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,7168,0.6099759737650553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,5120,0.14467467202080622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,7168,0.75400267706977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,6144,0.5268213484022353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,4096,0.12168622016906738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,6144,0.6505129072401259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,5120,0.4414053228166368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,3584,0.10679822497897679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,5120,0.5495120154486762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,4096,0.3559564484490289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,4096,0.44596264097425675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,3072,0.0971484449174669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,3584,0.39406755235460067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,3072,0.2706071005927192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,2560,0.08255644639333089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,3072,0.3430302143096924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,2560,0.22709690199957955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,2048,0.05788177914089627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,2560,0.2886524465348986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,3584,0.31326132350497776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,2048,0.17913511064317492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,1536,0.05052888724539015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,2048,0.2301351229349772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,1024,0.035466667678621076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,1536,0.13491911358303493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,1536,0.17499644226498076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,768,0.027172444595230952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,1024,0.12666577763027614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,1024,0.09613244401084052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,512,0.019313777486483257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,768,0.0767786635292901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,768,0.1028311120139228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,256,0.016183111402723525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,512,0.08046666781107585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,512,0.05865955352783203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,128,0.013919111755159167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,256,0.04233155647913615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,256,0.05608533488379585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,64,0.012085333466529846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3584,128,0.04184266593721178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,128,0.03717866539955139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3584,32,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,64,0.03329688972897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3584,32,0.03305777907371521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,65536,1.8799279530843098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,51200,1.3357475068834093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,65536,5.48856438530816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,16384,0.42371379004584414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,51200,4.435787624782987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,65536,6.512778811984592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,51200,5.098846011691623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,16384,1.3167653613620334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,16384,1.6455661985609267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,12288,0.9826968510945638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,10240,0.2787840101453993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,12288,0.329133325152927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,12288,1.2398079766167533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,8192,0.20156089464823404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,10240,0.8198524581061469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,7168,0.18846933046976724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,10240,1.039165284898546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,8192,0.6575359768337673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,6144,0.15819733672671848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,8192,0.8391813172234429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,7168,0.5800924301147461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,5120,0.13647288746303982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,7168,0.7376569112141927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,6144,0.4994959831237793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,4096,0.10239555438359578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,5120,0.41831466886732316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,6144,0.6379226578606499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,3584,0.08919288714726765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,5120,0.5358088811238607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,4096,0.3375111156039768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,3072,0.07699288924535115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,4096,0.43646756807963055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,3584,0.38483556111653644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,3584,0.2976293298933241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,3072,0.256751987669203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,2560,0.06743644343482123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,2048,0.05888444185256958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,3072,0.3348088794284397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,2560,0.21558666229248047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,2560,0.28253332773844403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,1536,0.04564355479346382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,2048,0.1692204475402832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,1024,0.031457778480317854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,2048,0.22471200095282662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,1536,0.12576533688439265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,768,0.024328889118300542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,1536,0.17131733894348145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,1024,0.09012977944480048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,512,0.019055111540688407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,1024,0.123325334654914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,768,0.07260533173878987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,768,0.10035644637213813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,512,0.07764799727333917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,256,0.052566223674350314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,256,0.039628442790773176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,128,0.013369777964221107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,3072,128,0.03879022267129686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,128,0.03319911162058512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,64,0.01147377739350001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,32,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,64,0.02991644541422526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,32,0.0292595558696323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,3072,512,0.054402665959464185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,65536,1.5007395214504664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,3072,256,0.01443288889196184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,51200,1.2010249031914604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,65536,5.2516326904296875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,51200,4.032024807400174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,16384,0.3721608850691054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,65536,6.252671135796441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,16384,1.241681734720866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,12288,0.25946399900648326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,51200,4.889662424723308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,12288,0.9291466606987847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,16384,1.5803733401828344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,10240,0.22576533423529732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,8192,0.1808862156338162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,12288,1.1939982308281791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,10240,0.7777466773986816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,8192,0.6234613524542915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,7168,0.17024978001912436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,10240,0.9987057579888238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,6144,0.13842311170366076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,8192,0.8074080149332682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,7168,0.5494968626234267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,6144,0.47281954023573136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,5120,0.12378311157226562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,5120,0.396854215198093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,6144,0.6131475766499838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,5120,0.5154817899068197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,4096,0.10164444314108954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,3584,0.0877679983774821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,4096,0.32102489471435547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,4096,0.4183777703179254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,3584,0.2813822163475884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,7168,0.7105466524759928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,3072,0.0768337779574924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,2560,0.06524799929724799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,3072,0.24314843283759224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,3072,0.3212249014112684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,2560,0.20311466852823892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,2048,0.05324000120162964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,2560,0.2691360049777561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,1536,0.04237511091762119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,2048,0.16005333264668783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,2048,0.21505510807037354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,1024,0.029271999994913738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,1536,0.1179857783847385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,1536,0.1632257832421197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,1024,0.08467111322614883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,768,0.022521777285469904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,1024,0.11731555726793076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,512,0.01753155555990007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,768,0.06807200113932292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,3584,0.37027645111083984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,256,0.01308444473478529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,512,0.05065422256787618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,512,0.07362933291329278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,128,0.011034666664070554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,256,0.0507315562831031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,256,0.03581422236230638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,64,0.009714666340086196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,128,0.029967112673653498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,128,0.03667111198107401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2560,32,0.0097884444726838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,32,0.02641511129008399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2560,64,0.02719555629624261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2560,768,0.0956382221645779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,65536,1.1281440522935655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,51200,0.8523582352532281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,51200,3.773749457465278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,65536,4.80545171101888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,16384,0.2948559919993083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,16384,1.1706355412801106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,51200,4.683328840467665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,12288,0.2192444403966268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,65536,5.990555657280816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,10240,0.1918097734451294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,16384,1.5181786219278972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,12288,0.876688904232449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,8192,0.14454844262864855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,10240,0.7337697876824273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,12288,1.144635518391927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,7168,0.13252888785468206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,10240,0.9600960413614908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,8192,0.5886622005038792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,6144,0.1089128918117947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,7168,0.5178124639723036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,8192,0.7744124200608996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,7168,0.6815013355678983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,5120,0.0882800022761027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,6144,0.4460284445020888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,4096,0.07194666730033027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,6144,0.5874257617526585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,5120,0.3738124370574951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,3584,0.06714222166273329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,5120,0.4951288965013292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,4096,0.30208712153964573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,3072,0.05723110834757487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,4096,0.4011840025583903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,3584,0.26524445745680064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,2560,0.04663199848598904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,3584,0.3551448980967204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,3072,0.22907911406622994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,2048,0.03979555434650845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,3072,0.30744889047410756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,2560,0.19043733014000785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,1536,0.030856887499491375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,2560,0.25826843579610187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,2048,0.15074400107065836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,2048,0.20505244202084014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,1024,0.021385777327749465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,1536,0.11152444945441352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,1536,0.1556604438357883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,768,0.017344888713624742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,1024,0.07729155487484403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,1024,0.11198222637176514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,512,0.013092444174819522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,768,0.09078044361538357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,512,0.045239110787709556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,256,0.010406222608354356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,512,0.06978488630718656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,256,0.03172711200184292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,256,0.04727377825313144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,128,0.00965244405799442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,128,0.02607822252644433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,2048,128,0.03518844313091702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,64,0.009017777939637503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,2048,32,0.009362666971153682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,64,0.02375288969940609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,32,0.023364444573720295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,2048,768,0.0633173320028517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,65536,0.8482177522447375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,51200,0.7322488890753852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,65536,4.405053456624349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,16384,0.2184986670811971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,51200,3.4446640014648438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,16384,1.094747543334961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,12288,0.1657902267244127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,65536,5.8557688395182295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,16384,1.4841377470228407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,10240,0.14601600170135498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,12288,0.8226684464348687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,12288,1.1187537511189778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,8192,0.11526666747199164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,10240,0.6884302033318414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,10240,0.9370595084296333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,7168,0.10452889071570502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,8192,0.5530915790134007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,8192,0.7566951115926107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,6144,0.08888177739249335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,51200,4.585169898139106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,7168,0.4863226678636339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,5120,0.07153599792056613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,7168,0.664588451385498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,6144,0.41967378722296816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,4096,0.056559999783833824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,5120,0.3513351016574436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,6144,0.5748622152540419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,5120,0.4827875561184353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,3584,0.053156445423762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,4096,0.2846382194095188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,3072,0.045448889334996544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,4096,0.3919866614871555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,3584,0.250270234213935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,2560,0.038731555143992104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,3584,0.34608978695339626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,3072,0.21623999542660186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,2048,0.030966222286224365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,3072,0.2995484405093723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,2560,0.18098666932847765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,1536,0.024523556232452393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,2048,0.14164177576700845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,2048,0.19982666439480254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,1024,0.018009778526094224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,1536,0.10444355673260158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,1536,0.15171199374728733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,1024,0.07219911283916898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,768,0.015067555838161044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,1024,0.10827200280295478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,512,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,768,0.053732444842656456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,768,0.08766133255428737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,256,0.009848000274764167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,512,0.04144800040456984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,256,0.028860443168216284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,128,0.009019555317031013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,256,0.04468444320890638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,2560,0.25181955761379665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,128,0.023405333360036213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,128,0.03353600038422479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,32,0.007607999775144789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,64,0.021024000313546922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1536,32,0.020607110526826646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1536,512,0.06680622365739611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1536,64,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,51200,0.4692391289605035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,65536,0.5591102176242405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,65536,4.079245249430339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,16384,0.14472088548872206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,51200,3.1970418294270835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,16384,1.02729246351454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,12288,0.10865333345201279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,65536,5.600813123914931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,51200,4.383761088053386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,16384,1.4200088712904189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,10240,0.09456711345248753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,12288,0.772797319624159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,8192,0.07455378108554415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,10240,0.6466479831271702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,12288,1.0710408952501085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,10240,0.8976187176174588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,7168,0.06842755609088473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,8192,0.5203030904134115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,7168,0.4572941992017958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,8192,0.7233555581834582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,6144,0.0595688886112637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,7168,0.6369928783840603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,5120,0.049287110567092896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,6144,0.3934320078955756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,6144,0.5498684247334799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,4096,0.04021155503061082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,5120,0.33130399386088055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,5120,0.4623369110955133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,3584,0.035897778140174016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,4096,0.26784443855285645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,3584,0.23515823152330187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,4096,0.3752693335215251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,3072,0.03115644388728672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,3584,0.3304693433973524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,2560,0.026312000221676294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,2560,0.1687742206785414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,3072,0.28641777568393284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,2048,0.02258488867017958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,2560,0.24053955078125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,1536,0.01778133379088508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,2048,0.13251377476586237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,2048,0.19017333454555937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,1024,0.013088888592190213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,1536,0.09746222363577949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,1536,0.14423733287387425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,768,0.010840888652536603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,1024,0.06653155883153279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,1024,0.1029040018717448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,512,0.0087333329849773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,768,0.04998666710323758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,768,0.08291377623875935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,256,0.006800888727108638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,512,0.03640177845954895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,512,0.060917331112755664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,3072,0.2034320036570231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,128,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,256,0.025763556361198425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,64,0.005888888819350137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,256,0.0418888893392351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,1024,128,0.031073778867721558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,1024,32,0.005617777920431561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,128,0.0206675562593672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,64,0.018207111292415194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,1024,32,0.017457778255144756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,65536,0.4358533223470052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,51200,0.3433369000752767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,51200,3.0840337541368275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,16384,0.11311288674672444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,65536,3.949040730794271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,16384,0.994018660651313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,51200,4.282048119439019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,12288,0.08415910932752822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,16384,1.3879067103068035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,10240,0.07691733042399089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,65536,5.476193745930989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,12288,0.7491057713826498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,8192,0.06266311142179701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,12288,1.0463831159803603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,10240,0.6270835664537218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,8192,0.5038204193115234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,7168,0.05496177739567227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,10240,0.8779039912753634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,6144,0.04907200071546766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,8192,0.7064879735310873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,7168,0.4427226649390326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,5120,0.04117244482040405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,6144,0.3822657797071669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,7168,0.6220151053534614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,4096,0.03226844469706217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,6144,0.5366044574313694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,5120,0.32061245706346303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,3584,0.02916533417171902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,5120,0.45194223192003036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,4096,0.259880887137519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,3584,0.22784177462259927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,4096,0.365706655714247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,3072,0.02532622218132019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,2560,0.021942221456103857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,3584,0.3234737714131673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,3072,0.19637510511610243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,2048,0.017886221408843994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,2560,0.1646933290693495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,2560,0.23540266354878744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,2048,0.12832978036668566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,1536,0.015152000718646579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,2048,0.18536444505055746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,1024,0.011011555790901184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,1536,0.09431288639704387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,1536,0.14050755235883924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,768,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,1024,0.099872887134552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,1024,0.06571022007200453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,512,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,768,0.05028177632225884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,768,0.08004533582263522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,512,0.034834666384591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,3072,0.2793448766072591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,256,0.006664888726340399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,512,0.05948978000217014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,128,0.006255110932721033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,256,0.024358222881952923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,256,0.041215999258889094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,64,0.005929777605666055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,128,0.018572444717089336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,768,32,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,64,0.01651733285850949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,768,32,0.01615111033121745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,65536,0.30482223298814565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,768,128,0.030413332912656996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,51200,0.24714665942721895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,65536,3.7928496466742625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,51200,2.969971550835503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,16384,0.07809244261847602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,16384,0.957673814561632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,65536,5.348399268256293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,12288,0.06432177623112996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,16384,1.3560497495863173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,12288,0.7214550971984863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,12288,1.0224427117241752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,10240,0.6033911175198025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,8192,0.0417484442392985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,10240,0.8564702139960395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,10240,0.05315999852286445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,7168,0.03703022334310744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,8192,0.4858853552076552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,51200,4.181290520562066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,6144,0.033302221033308245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,7168,0.4273511038886176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,8192,0.6904417673746744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,5120,0.027569777435726587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,7168,0.6081377665201823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,6144,0.3677511215209961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,4096,0.022680888573328655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,5120,0.3089004357655843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,4096,0.25005332628885907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,5120,0.44165510601467556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,3584,0.020072888996866014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,4096,0.35753867361280656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,3584,0.2198613352245755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,3072,0.01738666660255856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,3584,0.3149946795569526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,2560,0.015016888578732809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,3072,0.1893493334452311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,3072,0.27275554339090985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,2560,0.15782933764987522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,2048,0.012839111189047495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,2560,0.22962488068474662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,2048,0.12354577912224664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,1536,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,2048,0.1797919935650296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,6144,0.5255146556430393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,1024,0.008630221916569604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,1536,0.09039822551939224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,1536,0.13627021842532688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,768,0.007285333342022366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,1024,0.09684977928797404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,1024,0.06267466810014513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,512,0.006265777680608962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,768,0.04678222205903795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,256,0.0052284445199701525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,512,0.0573306679725647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,512,0.033371554480658636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,128,0.004632888982693355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,256,0.022712000542216833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,256,0.03946400019857619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,64,0.004874666531880696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,128,0.028840889533360798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,128,0.01758311192194621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,512,32,0.004943111290534337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,64,0.015480889214409722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,65536,0.20085955990685356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,512,32,0.015241778559154935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,51200,0.15871200296613905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,512,768,0.07850311199824016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,51200,2.8721485137939453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,16384,0.057314667436811656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,65536,3.6664691501193576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,51200,4.078953637017144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,16384,0.9288631015353732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,12288,0.04326577650176155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,16384,1.3225893444485135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,10240,0.03652711047066583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,65536,5.219295925564237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,12288,0.6998986668056912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,8192,0.03112444281578064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,12288,0.9979812833997938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,10240,0.5850088861253526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,7168,0.026727999250094097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,10240,0.8366550869411893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,8192,0.47118669086032444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,7168,0.4135928948720296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,6144,0.02144177754720052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,8192,0.6737937927246094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,6144,0.35667377048068577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,5120,0.014838222000333997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,6144,0.5118364228142632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,5120,0.29947733879089355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,4096,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,5120,0.43025954564412433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,4096,0.242104000515408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,3584,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,7168,0.5933084487915039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,3584,0.21305778291490343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,4096,0.34920088450113934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,3072,0.01016977760526869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,3584,0.307166232003106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,2560,0.009380444056457942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,3072,0.2655342155032688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,3072,0.18423022164238823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,2048,0.008069333102968004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,2560,0.15244711769951716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,2560,0.22284266683790419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,1536,0.0069751110341813825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,2048,0.1193644470638699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,1024,0.005842666659090254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,1536,0.08638755480448405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,2048,0.1749040020836724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,1536,0.13260355260637072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,768,0.005344888816277186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,1024,0.05961155229144626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,512,0.004583111239804162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,1024,0.09289066659079658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,512,0.031159109539455835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,768,0.07560977670881483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,256,0.003863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,512,0.05547822184032864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,128,0.003806222230195999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,256,0.0373324453830719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,256,0.02131822208563487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,128,0.015835555063353646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,64,0.00361333332127995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,256,128,0.027451554934183758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,256,32,0.003591999825504091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,64,0.013829333086808523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,32,0.01346933344999949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,65536,0.1737955543729994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,51200,0.13608533806271023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,256,768,0.04451733496454027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,51200,2.817728890313042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,16384,0.04955733484692044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,65536,3.600746578640408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,16384,0.9121209250556098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,12288,0.03533688849873013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,51200,4.074370490180121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,16384,1.3185848659939237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,10240,0.030180444320042927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,65536,5.201441023084852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,12288,0.6875848770141602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,8192,0.025388444463411968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,10240,0.5750986735026041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,12288,0.9961421754625109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,8192,0.46275467342800564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,7168,0.022153778208626643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,10240,0.8331084251403809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,6144,0.01369244439734353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,8192,0.6715706719292535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,7168,0.40686488151550293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,5120,0.010770666930410596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,6144,0.3503555456797282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,6144,0.5096506542629665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,5120,0.294140444861518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,4096,0.008978666530715095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,5120,0.4286479949951172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,3584,0.008280889027648503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,4096,0.3468782107035319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,4096,0.23833955658806694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,3584,0.20862578021155465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,3072,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,7168,0.5914488898383247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,3072,0.1798257827758789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,2560,0.007269333634111617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,3584,0.3053662247127957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,2048,0.006241777704821692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,3072,0.2645004325442844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,2560,0.14957955148484972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,1536,0.005219555563396878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,2048,0.11667199929555257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,2560,0.222692436642117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,1024,0.004585777719815572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,2048,0.17461155520545113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,1536,0.08422578043407863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,1536,0.1313342253367106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,1024,0.057962669266594775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,768,0.04282844397756788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,1024,0.09277155664232041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,512,0.003899555653333664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,768,0.07512978050443861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,512,0.05485511157247755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,256,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,512,0.030717333157857258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,128,0.0032542221662071017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,256,0.020548444655206468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,256,0.0373306671778361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,128,0.015286222100257874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,64,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,768,0.004187555362780889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,128,32,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,64,0.013163555827405719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,128,32,0.01275288893116845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,65536,0.17006044917636445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,51200,0.1323502196205987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,16384,0.05080533358785841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,51200,2.7973289489746094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,12288,0.0343537794219123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,65536,3.578210618760851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,10240,0.02962488929430644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,12288,0.6823360125223795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,16384,0.9058977762858073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,10240,0.5704204241434733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,7168,0.02159999973244137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,7168,0.40212000740898984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,6144,0.011676444775528379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,8192,0.45923556221856016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,5120,0.009658666948477427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,8192,0.024580443898836773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,6144,0.34666045506795246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,4096,0.008288000192907121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,3584,0.007646222081449296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,4096,0.23505422804090711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,3072,0.006928000185224745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,3584,0.20673599508073595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,2560,0.006999111009968652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,5120,0.2919110986921522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,2048,0.006392000036107169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,3072,0.17794488535987008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,1536,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,2560,0.14788711071014404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,1024,0.004880889008442561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,1536,0.08344799942440456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,768,0.004344888859324985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,1024,0.05723110834757487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,2048,0.115466660923428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,512,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,256,0.0035662220584021676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,768,0.04192622171507942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,128,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,256,0.01997244523631202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,64,0.003142222141226133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,128,0.01443911095460256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,64,32,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,512,0.03035911255412632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,64,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,64,32,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,65536,0.16671821806165907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,51200,0.13303555382622614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,16384,0.05184444453981188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,16384,0.9239475462171766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,12288,0.03289599882231818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,12288,0.6962782012091743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,10240,0.02826666831970215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,51200,2.825318230523003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,10240,0.5702142185635036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,8192,0.023550222317377727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,7168,0.02068000038464864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,65536,3.5860328674316406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,6144,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,7168,0.40314756499396426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,5120,0.009349333743254343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,6144,0.34702489111158585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,4096,0.007375111182530721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,4096,0.23530756102667916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,5120,0.2922000090281169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,3584,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,3072,0.006675555474228329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,8192,0.4578515688578288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,2560,0.005666666560702854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,3072,0.17974755499098036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,2560,0.14925421608818903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,2048,0.0052728888889153796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,3584,0.20927111307779947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,2048,0.11657333374023438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,1024,0.004269333349333869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,1536,0.0853653351465861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,1024,0.05724889039993286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,768,0.003991999973853429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,512,0.0036675557494163513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,768,0.041834665669335254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,256,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,512,0.03058755397796631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,128,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,1536,0.005521777603361342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,64,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,128,0.014811555544535318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,64,0.012691555751694573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8192,32,32,0.002942222274012036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,32,0.012404444317022959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8192,32,256,0.019926221834288705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,16384,4.624274783664279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,16384,6.3301281399197045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8192,128,128,0.027053332991070215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,51200,14.452336629231771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,12288,3.520608054267036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,10240,3.034864001803928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,12288,4.396095275878906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,10240,3.5652364095052085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,8192,2.291269302368164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,8192,2.7482532925075955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,51200,22.298195732964412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,7168,2.0048906538221574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,6144,1.7367075814141169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,7168,2.3969874911838107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,5120,1.4096559948391383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,6144,2.024343066745334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,4096,1.1681067148844402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,5120,1.6904018190171983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,3584,0.963199085659451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,4096,1.3380568822224934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,3072,0.8253262307908801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,3584,1.1731253729926217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,2560,0.6958355373806424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,3072,1.0082826614379883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,2560,0.8537439770168729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,2048,0.5725866423712836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,1536,0.4045848846435547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,2048,0.6962960031297473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,1024,0.29219555854797363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,768,0.23256089952256945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,1536,0.5494177606370714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,1024,0.4190959930419922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,512,0.17113599512312147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,256,0.13450933827294245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,768,0.35175289048088926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,128,0.11155822541978623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,512,0.2770631048414442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,256,0.24800178739759657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,64,0.09514222542444865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,65536,32,0.10090488857693142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,64,0.238846222559611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,128,0.23527910974290636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,65536,32,0.2451662222544352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,51200,11.166747199164497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,65536,14.318115234375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,16384,3.676219516330295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,12288,2.690962685479058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,51200,17.76373969184028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,16384,5.061041937934028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,10240,2.293914582994249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,12288,3.5038454267713757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,8192,1.7526648839314778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,10240,2.8625395033094616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,65536,22.718858506944443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,7168,1.5212382210625544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,8192,2.186788347032335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,6144,1.432037353515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,7168,1.9181323581271703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,5120,1.0888755586412218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,4096,0.9122613271077474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,6144,1.6252382066514757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,5120,1.354839113023546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,3584,0.7516711023118762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,3072,0.669000890519884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,4096,1.0747750600179036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,3584,0.9368559519449869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,2560,0.5816871325174967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,3072,0.806421332889133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,2048,0.4500311215718587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,2560,0.6864035394456652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,1536,0.3366408877902561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,1536,0.440665774875217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,1024,0.2357111242082384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,2048,0.6244168811374241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,768,0.17894577980041504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,1024,0.3308266798655192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,512,0.14150133397844103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,256,0.10473510954115127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,128,0.08767555819617377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,512,0.22168889310624865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,768,0.26093512111239964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,256,0.19741243786282012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,64,0.07554577456580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,51200,32,0.07858400212393866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,128,0.18655644522772896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,64,0.1902079979578654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,51200,32,0.1951120032204522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,51200,3.498753865559896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,65536,4.486663818359375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,16384,1.1405555937025282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,12288,0.8562737570868598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,16384,1.905463112725152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,51200,6.682850731743707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,10240,0.7408000098334419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,12288,1.3493661880493164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,8192,0.64101775487264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,10240,1.1037075254652235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,65536,8.24247572157118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,8192,0.8639501995510526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,7168,0.5476604567633735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,6144,0.4621404541863336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,5120,0.37542221281263566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,7168,0.7593919965955945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,4096,0.3047955565982395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,6144,0.649869336022271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,5120,0.5403848754035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,3584,0.25595466295878094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,4096,0.4254986710018582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,3072,0.22302577230665419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,2560,0.1963813304901123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,3072,0.3190266556209988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,2048,0.1460053390926785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,2560,0.28746843338012695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,1536,0.11887466907501221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,2048,0.22162755330403647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,1024,0.07876977655622694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,1536,0.17467466990152994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,1024,0.1311617824766371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,768,0.06294933292600843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,512,0.04744799931844076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,3584,0.3717644479539659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,256,0.03857422206136916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,768,0.1044844455189175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,128,0.03122044603029887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,256,0.0753600001335144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,128,0.06999911202324761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,64,0.026521777113278706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,16384,32,0.028211557202868994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,64,0.0703751113679674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,32,0.07201155689027575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,16384,512,0.0870435569021437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,51200,2.675000932481554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,65536,3.3106248643663196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,16384,0.8629848692152234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,12288,0.6332897610134548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,16384,1.540721787346734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,10240,0.5296159850226508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,12288,1.0957057740953233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,51200,5.369698842366536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,8192,0.43195467525058323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,10240,0.9119449191623263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,7168,0.39810933007134336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,65536,6.670427534315322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,6144,0.3302737871805827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,7168,0.6222106615702311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,5120,0.27304445372687447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,4096,0.21106221940782335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,5120,0.4438275496164958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,6144,0.5343102349175347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,3584,0.18720622857411703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,4096,0.35054487652248806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,3072,0.16204978360070124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,2560,0.13577866554260254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,3584,0.30484533309936523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,3072,0.26126133071051705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,2048,0.11012533638212417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,2560,0.22220888402726915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,2048,0.18200177616543242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,1024,0.05919910801781548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,1536,0.1438773340649075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,768,0.04860977662934197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,8192,0.7083439826965332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,1024,0.10761511325836182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,512,0.03602755400869582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,256,0.027580445011456806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,768,0.08588000138600667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,128,0.02385155525472429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,512,0.07148177756203546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,64,0.020624000165197585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,256,0.06042844719356961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,32,0.021722666091389124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,128,0.05580266979005602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,64,0.05638666947682699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,12288,1536,0.08797066741519505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,12288,32,0.057328886455959745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,51200,2.2020026312934027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,65536,2.769650565253364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,16384,0.7118142445882162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,12288,0.5545840263366699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,16384,1.35627195570204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,10240,0.4502906799316406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,12288,0.9692861768934461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,51200,4.724733140733507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,8192,0.358749336666531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,10240,0.86554045147366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,8192,0.6299475563897027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,7168,0.3230648835500081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,65536,5.86740960015191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,6144,0.268958224190606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,5120,0.235555542839898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,7168,0.5543600188361274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,6144,0.47598308987087673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,4096,0.1799688869052463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,5120,0.39565155241224503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,3584,0.1647919946246677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,3072,0.13745066854688856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,4096,0.3128160105811225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,2560,0.11438577704959446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,3584,0.2708497842152913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,2048,0.09572711255815293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,1536,0.07872533135943942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,2560,0.19775466124216715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,3072,0.2336826589372423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,2048,0.16208799680074057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,768,0.04276622335116068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,1536,0.12889777289496526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,768,0.07687999804814656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,1024,0.09609778059853448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,512,0.03104089034928216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,256,0.023550222317377727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,256,0.05332444442643059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,512,0.06367733081181844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,128,0.020301333732075162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,64,0.017823111679818895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,1024,0.051536887884140015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,64,0.0492159989145067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,128,0.04909866717126635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,10240,32,0.018533334136009216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,10240,32,0.04993866549597847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,51200,1.8411956363254125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,65536,2.2623812357584634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,16384,0.6070879830254449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,12288,0.4599324332343207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,16384,1.1733777787950304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,10240,0.3804062207539876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,12288,0.8429910871717664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,8192,0.2926151222652859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,10240,0.7053928905063205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,8192,0.551228470272488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,7168,0.2542622354295519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,65536,5.202142079671224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,6144,0.2200853294796414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,5120,0.1918106608920627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,7168,0.48679468366834855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,6144,0.4174035655127631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,4096,0.14245866404639349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,51200,4.060516357421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,3584,0.12706666522555882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,4096,0.273747550116645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,3072,0.10924444595972697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,5120,0.34835465749104816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,2560,0.09460888968573676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,3584,0.23831733067830405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,2048,0.07515555620193481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,1536,0.0585742195447286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,3072,0.20449154906802705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,2048,0.14173332850138345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,1024,0.041899555259280734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,2560,0.1740408870908949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,768,0.0329075555006663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,512,0.024268445041444566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,1536,0.11237600114610459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,256,0.019090667366981506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,1024,0.08338933520846897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,128,0.016999999682108562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,512,0.05511733227305942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,768,0.06785421901279025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,64,0.014769777655601501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,8192,32,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,128,0.04041066765785217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,32,0.039529777235454984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,64,0.0406257775094774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,8192,256,0.045518222782346934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,51200,1.6118346320258246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,65536,2.084467569986979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,16384,0.5237661997477213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,12288,0.3861288759443495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,16384,1.0862141715155709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,10240,0.3186986711290148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,12288,0.7801564534505209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,8192,0.28028978241814506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,10240,0.6705679893493652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,7168,0.22249510553148058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,51200,3.670750935872396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,8192,0.5305448638068305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,65536,4.79188707139757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,6144,0.18893333276112875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,7168,0.4519351323445638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,5120,0.1579386658138699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,4096,0.13502489195929632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,3584,0.11146756013234456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,5120,0.32496711942884654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,6144,0.38961511188083225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,3072,0.09982666704389785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,4096,0.2548648781246609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,2560,0.08076266447703044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,3584,0.22089155515034994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,2048,0.06985333230760363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,1536,0.05533777674039205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,3072,0.1905991103914049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,2048,0.13208799892001682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,1024,0.03891822364595201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,768,0.029903110530641343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,1536,0.10420088635550605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,512,0.020968000094095867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,1024,0.07804444763395521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,256,0.017679999272028606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,512,0.05102488729688856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,128,0.015150222513410779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,768,0.06313421991136339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,256,0.04051733348104689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,64,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,128,0.03671377897262573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,64,0.034712000025643244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,32,0.03464355402522617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,7168,2560,0.16112266646491155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,7168,32,0.014254222313563028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,65536,1.7692293590969508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,51200,1.2865751054551866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,16384,0.44403910636901855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,12288,0.334150234858195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,16384,0.9909991158379449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,65536,4.272254096137153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,10240,0.27875733375549316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,12288,0.7191911273532443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,8192,0.2265226576063368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,7168,0.19392177793714735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,51200,3.4131857554117837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,10240,0.5954222149319118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,8192,0.4745350943671332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,6144,0.16450044843885633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,5120,0.13737066586812338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,6144,0.3600160015953912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,4096,0.11027466588550144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,7168,0.42046666145324707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,3584,0.09582933452394272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,5120,0.300111108356052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,3072,0.08333688974380493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,2560,0.0732133322291904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,4096,0.23634044329325357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,2048,0.05997777647442288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,3072,0.17461600568559435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,1536,0.04570133487383524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,3584,0.20431555642022026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,2560,0.14972799354129368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,1024,0.03320711188846164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,2048,0.12258133623335098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,1536,0.09644089142481486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,512,0.01927911076280806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,1024,0.07207733392715454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,256,0.015612444943851896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,512,0.045893334680133395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,128,0.015843555331230164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,256,0.03588355580965678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,64,0.01389955480893453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,128,0.0325911111301846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,32,0.014497778481907315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,64,0.03119644522666931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,32,0.030842665168974135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,6144,768,0.02570844524436527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,6144,768,0.057948443624708385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,65536,1.4529359605577257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,51200,1.1229804356892903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,16384,0.3779004414876302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,12288,0.30023733774820965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,16384,0.9047697914971246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,10240,0.2392044332292345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,12288,0.6764959759182401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,8192,0.1844764418072171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,10240,0.5475173526340061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,51200,3.055511898464627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,7168,0.16420000129275852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,65536,3.934409671359592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,6144,0.1418559948603312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,7168,0.38451290130615234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,5120,0.11132089296976726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,4096,0.09295911259121364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,6144,0.33039821518792045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,5120,0.2764577865600586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,3584,0.08225866821077135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,3072,0.0705911119778951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,4096,0.2170177830590142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,8192,0.43617065747578937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,2560,0.059642665916019015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,3584,0.1877511077457004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,2048,0.05158222383923001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,3072,0.1598222255706787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,1536,0.04035377833578322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,1024,0.027994665834638808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,2560,0.13570488823784724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,768,0.021535111798180476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,1536,0.08863644467459784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,512,0.01626488897535536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,1024,0.0659244457880656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,768,0.05210577779346042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,512,0.04034488730960422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,256,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,128,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,64,0.010023111270533668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,256,0.031502223677105375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,5120,32,0.01071999967098236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,128,0.02863911125395033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,64,0.027816000911924574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,32,0.027476444840431213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,5120,2048,0.11239022678799099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,51200,0.8851440217759875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,65536,1.097608036465115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,16384,0.3064720100826687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,12288,0.22699644830491808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,16384,0.7985475328233508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,10240,0.1842399968041314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,12288,0.6234684520297581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,8192,0.1500097778108385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,10240,0.49161243438720703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,8192,0.3961173428429498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,51200,2.681009716457791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,7168,0.12883467144436306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,6144,0.10965689023335774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,7168,0.34943556785583496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,65536,3.5106809404161243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,5120,0.09744799799389309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,4096,0.07472799883948432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,6144,0.3016471068064372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,5120,0.2510293324788411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,3584,0.0688577757941352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,3072,0.0570160018073188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,4096,0.19812444845835367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,2560,0.048398223188188344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,3584,0.1709831158320109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,2048,0.04131022095680237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,2560,0.12352711624569362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,3072,0.146978669696384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,1536,0.031585776143603854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,2048,0.10044266780217488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,1024,0.022816888160175745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,768,0.018421333697107103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,1024,0.05846044752332899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,1536,0.0812879999478658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,512,0.013930666777822705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,768,0.046478221813837685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,256,0.02753955622514089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,512,0.03599822190072801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,128,0.009461333354314169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,64,0.008654221892356873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,128,0.025440000825458105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,64,0.02443555494149526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,32,0.008984000318580205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,4096,32,0.024117334021462336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,4096,256,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,65536,0.970211558871799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,16384,0.25609511799282497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,51200,0.7988950941297742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,12288,0.20779732863108316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,16384,0.7516524526807996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,12288,0.5824622048272027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,10240,0.1667475567923652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,8192,0.13570933871799046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,10240,0.46907912360297305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,7168,0.11443644099765354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,51200,2.497054206000434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,8192,0.3773350980546739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,6144,0.10494755374060737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,5120,0.08322933647367689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,7168,0.33287021848890513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,65536,3.2754222022162547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,4096,0.07404266463385688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,3584,0.06343911091486613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,6144,0.28741955757141113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,5120,0.23965332243177626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,4096,0.18836355209350586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,3072,0.05561155743069119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,2560,0.04599822229809231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,3584,0.16259822580549452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,2048,0.03974755605061849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,3072,0.13869778315226236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,1536,0.02929511004024082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,1024,0.022523555490705702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,2048,0.09538844558927749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,2560,0.11722755432128906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,768,0.017903111047214933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,1536,0.07600977685716417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,512,0.015110222829712762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,256,0.011695110963450538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,1024,0.054802666107813515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,128,0.008752000000741746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,512,0.03344444433848063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,768,0.04464088877042135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,256,0.025763556361198425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,128,0.023523555861579046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,32,0.007954667011896769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,32,0.022727999422285292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3584,64,0.023129777775870428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,65536,0.8502622180514865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,51200,0.6487173504299587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3584,64,0.007625777688291337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,16384,0.23214756117926705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,12288,0.17210755083296034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,16384,0.7063440216912164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,10240,0.13703644275665283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,12288,0.532582230038113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,8192,0.11667199929555257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,10240,0.44507111443413633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,7168,0.09702933496899074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,8192,0.35889511638217503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,6144,0.0837564468383789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,7168,0.31611288918389213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,51200,2.369464874267578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,65536,3.0479884677463107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,5120,0.06937244203355578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,4096,0.05654311180114746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,3584,0.05129244592454699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,6144,0.2725893391503228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,4096,0.17866755856408012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,5120,0.2270053227742513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,2560,0.038831998904546104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,3072,0.04572711057133145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,3584,0.15407110585106745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,2048,0.0318213336997562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,3072,0.13127022319369847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,1536,0.02404533326625824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,2560,0.11046133438746135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,1024,0.018256889449225534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,2048,0.09009866582022773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,768,0.01482488877243466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,1536,0.06943733162350126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,512,0.010688888529936472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,1024,0.05120889014667935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,256,0.008971555365456475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,768,0.040786667002571955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,512,0.031214220656288996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,128,0.008045333127180735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,256,0.024056888288921777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,64,0.007277333074145847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,3072,32,0.007354666789372762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,128,0.021958221991856892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,64,0.021233777205149334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,3072,32,0.021003555920388963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,65536,0.7743448681301541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,51200,0.6064560148451064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,16384,0.19989066653781465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,16384,0.6741884549458822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,12288,0.15312443839179143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,12288,0.5029208925035265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,10240,0.1445502175225152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,65536,2.805760065714518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,8192,0.11515822675493027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,51200,2.1553448571099176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,7168,0.0974853303697374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,10240,0.42112000783284503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,6144,0.08467111322614883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,8192,0.33939290046691895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,7168,0.29918400446573895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,6144,0.25789422459072536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,4096,0.0599235561158922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,3584,0.05433333251211378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,5120,0.21536178059048125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,3072,0.0458808872434828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,4096,0.1696479982799954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,2560,0.03384266628159417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,3072,0.12441866927676731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,2048,0.026871111657884385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,2560,0.10481422477298313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,1536,0.02293155590693156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,2048,0.085207998752594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,1024,0.015797333584891427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,1536,0.06528533167309232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,5120,0.0735679997338189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,768,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,512,0.010031999813185798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,1024,0.04775289032194349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,256,0.008354666332403818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,768,0.0384497774971856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,512,0.029504888587527808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,128,0.0079644446571668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,256,0.022485334012243483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,3584,0.14623822106255427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,128,0.020277332928445604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,32,0.007252444823582967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,64,0.01960177719593048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2560,32,0.019296889503796894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,65536,0.5549475351969401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,51200,0.44122666782803005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,16384,0.14716711309221056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2560,64,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,12288,0.10943288935555352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,16384,0.6372026867336696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,10240,0.09732800059848362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,12288,0.4758586883544922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,8192,0.08146578073501587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,10240,0.39839911460876465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,7168,0.06829777691099378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,8192,0.3214844332800971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,6144,0.06037955813937717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,65536,2.5370943281385636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,51200,1.9871919420030382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,5120,0.048307554589377515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,7168,0.2830391195085314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,4096,0.03970400161213345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,3584,0.03527288966708713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,6144,0.24464888042873809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,5120,0.20267111725277373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,4096,0.16005778312683105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,3072,0.031247112486097548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,2560,0.028012444575627644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,3584,0.1381022267871433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,2048,0.02335111134582096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,1536,0.018056000272432964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,3072,0.11702133549584283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,1024,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,2560,0.09908355606926812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,2048,0.08046933015187581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,768,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,1536,0.06093600061204699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,512,0.008773333496517604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,1024,0.04302933481004503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,256,0.006976889239417181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,768,0.03604977660708957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,128,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,512,0.027806222438812256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,256,0.021327111456129286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,64,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,2048,32,0.006005333529578314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,128,0.018600889378123812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,64,0.01830222209294637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,2048,32,0.017839110559887357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,65536,0.4273413287268744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,51200,0.33506043752034503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,16384,0.11070133580101861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,16384,0.5922755665249294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,12288,0.08452444606357151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,12288,0.4465386602613661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,10240,0.07895200120078193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,51200,1.8398693932427301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,65536,2.3424434661865234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,8192,0.06132088767157661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,7168,0.05857866340213352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,10240,0.37430400318569607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,6144,0.04672888914744059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,8192,0.3021591239505344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,7168,0.2660106552971734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,5120,0.03774311145146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,4096,0.030230220821168687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,6144,0.2288435565100776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,3584,0.027884443600972492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,5120,0.19104088677300346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,4096,0.15032445059882268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,3072,0.024490666058328416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,2560,0.020639111598332722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,3584,0.1296506722768148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,2048,0.01755199995305803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,3072,0.10972088575363159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,2560,0.09186044666502212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,1536,0.014095111025704278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,1024,0.010663111176755695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,2048,0.07491911119884916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,1536,0.058837334314982094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,768,0.008938666847017076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,1024,0.041202667686674334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,512,0.007308444215191736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,256,0.006018666757477655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,768,0.03288177649180094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,128,0.005391110976537068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,256,0.019646222392718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,64,0.00500177757607566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,128,0.01687022215790219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1536,32,0.0052435555391841466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,64,0.016608888904253643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,32,0.016232000456915963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,65536,0.30159110493130153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,51200,0.2337555620405409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1536,512,0.02583822276857164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,16384,0.0774657792515225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,12288,0.05851111147138807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,16384,0.5565742386711968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,10240,0.049950222174326576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,12288,0.4212808873918321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,8192,0.04258133305443657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,10240,0.35314488410949707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,7168,0.038190222448772855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,51200,1.7143466737535265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,65536,2.1908719804551864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,6144,0.03153866529464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,5120,0.026742221580611333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,7168,0.25119733810424805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,4096,0.021941334009170532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,6144,0.2158435583114624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,3584,0.020572443803151447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,5120,0.1800311141543918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,3072,0.018231110440360177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,4096,0.14185421996646458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,3584,0.12257956133948432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,3072,0.10358933607737224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,2048,0.01389955480893453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,2560,0.08653244707319473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,1536,0.0103502223889033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,2048,0.07113866673575507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,1024,0.008403555386596257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,8192,0.28563110033671063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,1536,0.05445955528153313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,768,0.007594666547245449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,512,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,1024,0.037427554527918495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,256,0.0053342220683892565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,768,0.030572444200515747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,128,0.004540444248252445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,512,0.02437066700723436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,64,0.004266666869322459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,256,0.01793155570824941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,32,0.004587555511130227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,128,0.015487111277050443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,1024,2560,0.015172445111804538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,32,0.014503111441930136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,65536,0.22820178667704263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,1024,64,0.015235554840829639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,51200,0.1860577795240614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,16384,0.06154133213890923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,16384,0.5380755530463325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,12288,0.04413333204057482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,12288,0.40697420967949766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,10240,0.03831200136078729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,51200,1.652101304796007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,8192,0.0317555566628774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,65536,2.1113103230794272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,10240,0.3413902123769124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,7168,0.026198221577538386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,6144,0.022691556149058874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,8192,0.27572443750169545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,5120,0.01945688823858897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,7168,0.24248178799947104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,4096,0.016502221425374348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,6144,0.2089475525750054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,3584,0.014829332629839579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,5120,0.1744542254341973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,3072,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,4096,0.1373173395792643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,3584,0.11841066678365071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,2560,0.01185155577129788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,2048,0.010311110979980892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,3072,0.09997777806388007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,1536,0.008440888590282863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,2560,0.08367733160654704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,1024,0.006948444578382704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,1536,0.05196266704135471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,768,0.0062871111763848206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,1024,0.03601600064171685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,512,0.005423111220200856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,768,0.029476443926493328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,256,0.004581333448489507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,512,0.023352000448438857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,128,0.0041955556306574075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,256,0.017208000024159748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,64,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,128,0.01481599940194024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,768,32,0.003939555750952827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,64,0.014149333039919535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,32,0.013779555757840475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,65536,0.17526488833957246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,768,2048,0.06783466868930392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,51200,0.13778666655222574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,16384,0.04715377754635281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,16384,0.5200515323215061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,12288,0.0365226666132609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,12288,0.39162932501898867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,10240,0.03017510970433553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,51200,1.5911280314127605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,8192,0.022589332527584497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,7168,0.020644444558355544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,10240,0.3291369014316135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,65536,2.0326336754692926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,6144,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,5120,0.015452444553375244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,8192,0.2659671041700575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,4096,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,7168,0.23435645633273652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,6144,0.2013173368242052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,3584,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,5120,0.16843999756707084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,3072,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,4096,0.13245333565606013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,2560,0.009689778089523315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,2048,0.0083155557513237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,3584,0.11344799730512832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,1536,0.007262222468852997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,3072,0.09592177470525105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,1024,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,2560,0.080077330271403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,768,0.005612444546487596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,2048,0.06574133369657728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,1536,0.049294223388036094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,1024,0.03470044334729513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,512,0.004902222090297275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,256,0.004211555752489301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,128,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,768,0.028194665908813477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,512,0.022409778502252366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,64,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,512,32,0.003863999827040566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,256,0.016570667425791424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,128,0.014173333843549093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,64,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,65536,0.1203742159737481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,512,32,0.01313688854376475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,51200,0.09714221954345703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,16384,0.034084445900387235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,12288,0.02718399961789449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,12288,0.3781760003831651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,16384,0.4995457861158583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,10240,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,8192,0.018473777506086562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,10240,0.3173697789510091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,7168,0.016172443826993305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,8192,0.2559715641869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,6144,0.014841778410805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,7168,0.22548532485961914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,5120,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,6144,0.19423645072513154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,65536,1.9535609351264105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,5120,0.16200978226131862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,4096,0.01089599976936976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,3584,0.009741333623727163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,4096,0.12722844547695583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,3584,0.10982044537862141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,3072,0.008299555215570662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,2560,0.007344000041484833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,3072,0.0920906662940979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,2560,0.07667555411656697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,2048,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,1536,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,51200,1.5318782594468858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,1024,0.004951999833186467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,2048,0.062330669826931424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,1536,0.04669333497683207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,1024,0.03327111072010464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,768,0.02695911129315694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,512,0.004039999925427967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,256,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,512,0.02131822208563487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,128,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,256,0.01570311188697815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,64,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,128,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,32,0.0032924444725116095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,64,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,256,32,0.012489777472284106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,256,768,0.004549333204825719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,65536,0.09761510954962836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,51200,0.076928887102339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,16384,0.033702221181657575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,12288,0.02661866611904568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,16384,0.49021599027845597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,10240,0.01513155632548862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,12288,0.3708648946550157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,8192,0.013053333593739403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,51200,1.502854241265191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,10240,0.3110577795240614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,7168,0.011476444701353708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,6144,0.01072266697883606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,65536,1.9176133473714192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,5120,0.009335999687512716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,8192,0.2512613402472602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,4096,0.007997333175606197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,7168,0.22079732682969835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,3584,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,6144,0.1908808946609497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,5120,0.15937777360280356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,3072,0.006633777585294511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,3584,0.10675022337171768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,2560,0.006287999865081575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,2048,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,3072,0.09015822410583496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,1536,0.004932444542646408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,2560,0.07507999738057454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,1024,0.004361777669853634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,2048,0.06096711423661974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,1536,0.04468977782461378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,1024,0.031522668070263334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,512,0.003510222252872255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,768,0.026392888691690233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,256,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,512,0.02092444399992625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,256,0.015126221709781222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,128,0.012712889247470431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,4096,0.12541866302490234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,32,0.00309955556359556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,64,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,65536,0.09573244386249119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,128,32,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,51200,0.07494844330681695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,16384,0.0291920006275177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,128,768,0.004192000048028098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,12288,0.020939555433061387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,16384,0.49013598759969074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,10240,0.013287111288971372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,12288,0.37095199690924746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,8192,0.01219911128282547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,10240,0.31068534321255153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,7168,0.010666666759385003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,51200,1.5019333097669814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,6144,0.010022222167915767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,7168,0.22070488664839003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,5120,0.008399110701349046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,65536,1.917085435655382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,4096,0.007605333295133378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,6144,0.19083466794755724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,3584,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,5120,0.15854399734073216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,3072,0.006244444598754247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,4096,0.12403999434577094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,2560,0.005622222191757626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,3584,0.10660266876220703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,2048,0.005447111195988125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,3072,0.0894462201330397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,1536,0.004602666530344221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,2560,0.07501777675416735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,2048,0.06112800041834513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,1536,0.04526844289567736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,1024,0.0041680000722408295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,768,0.003942222230964237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,8192,0.2512693405151367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,512,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,1024,0.03131466772821214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,256,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,768,0.02633511192268795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,128,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,512,0.021041777398851182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,64,0.003015999992688497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,128,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,64,32,0.0029528888149393927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,64,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,32,0.011689778003427716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,65536,0.09404177798165216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,51200,0.07503110832638211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,64,256,0.0148044443792767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,16384,0.02725511127048069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,16384,0.49001068539089626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,12288,0.017874666386180453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,51200,1.5022506713867188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,65536,1.9205527835422094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,10240,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,8192,0.01054044481780794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,12288,0.3708995448218451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,7168,0.009535111486911774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,6144,0.009353777600659264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,10240,0.3107697698805067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,7168,0.22124177879757354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,8192,0.2512284384833442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,5120,0.011881777809725868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,4096,0.010047111246320937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,6144,0.19020356072319877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,3584,0.009399111072222391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,3072,0.008365333080291748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,5120,0.15857156117757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,4096,0.12495377328660752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,2560,0.007287110719415877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,3584,0.10770666599273682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,2048,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,1536,0.004625777817434735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,3072,0.0903795560201009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,2560,0.0761422250005934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,2048,0.060659554269578725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,1024,0.003973333372010125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,768,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,1536,0.045350223779678345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,512,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,1024,0.031565331750445895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,768,0.02645955483118693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,128,0.003160888950030009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,512,0.02091288897726271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,64,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,256,0.014497778481907315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4353,32,32,0.003148444410827425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,128,0.012499555945396423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,32,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4353,32,64,0.012039999994966718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,16384,4.202097574869792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,16384,6.2644399007161455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,51200,13.179380628797743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,12288,3.1892852783203125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,16384,4.726830376519097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,12288,4.307017856174045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,51200,16.02435811360677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,51200,21.22912089029948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,12288,3.487255096435547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,10240,2.873607211642795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,10240,3.511553870307075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,10240,2.881345748901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,8192,2.60445425245497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,8192,2.6970969306098094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,7168,2.0593128204345703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,8192,2.3073103162977433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,7168,2.3600845336914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,6144,1.705016877916124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,7168,2.03312259250217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,6144,1.9947200351291234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,6144,1.766635576883952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,5120,1.5022675196329753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,5120,1.6701306237114801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,4096,1.1759938134087455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,5120,1.487860467698839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,4096,1.3227181964450412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,4096,1.1964569091796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,3584,0.9645182291666666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,3584,1.153448846605089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,3072,0.943088001675076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,3584,1.0517501831054688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,3072,0.9933475918240018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,3072,0.9145662519666883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,2560,0.7632622188991971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,2560,0.8469066619873047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,2048,0.5534488889906142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,2560,0.7764079835679797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,2048,0.6909884346856011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,1536,0.42966755231221515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,2048,0.6473537551032172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,1536,0.5435653262668186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,1024,0.29831110106574166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,1536,0.514329751332601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,1024,0.40886489550272626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,768,0.23592710494995117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,1024,0.3902204566531711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,768,0.32291822963290745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,768,0.34601955943637425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,512,0.1739351087146335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,256,0.12511111630333796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,512,0.2740817864735921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,512,0.25740710894266766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,128,0.10535911056730483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,256,0.24585511949327257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,256,0.20082578394148085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,64,0.0886240005493164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,128,0.2316488954755995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,65536,32,0.09287555350197686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,64,0.23582845264010957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,65536,32,0.2425128883785672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,65536,128,0.15592443943023682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,51200,11.160650465223524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,65536,14.349872165256075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,65536,16.537167019314236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,65536,22.264851888020832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,16384,3.6437208387586804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,16384,4.974502987331815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,51200,16.913279215494793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,12288,2.6288301679823136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,51200,12.176349216037325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,16384,3.8591147528754344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,10240,2.2933404710557723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,12288,3.458855946858724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,10240,2.8051333957248263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,12288,2.8073298136393228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,8192,1.6740230984157984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,8192,2.15882937113444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,10240,2.3168036142985025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,7168,1.4305173026190863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,8192,1.876950158013238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,6144,1.2207315233018663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,7168,1.8889545864529078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,7168,1.6339022318522136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,6144,1.6053813298543294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,5120,0.973274654812283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,4096,0.8143911361694336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,6144,1.4162222544352214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,5120,1.3390248616536458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,5120,1.1949191623263888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,3584,0.6925137837727865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,4096,1.0558320151435006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,4096,0.9671804640028211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,3584,0.9274560080634223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,3584,0.8514729075961642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,3072,0.5852559937371148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,3072,0.7958595487806531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,2560,0.5323368708292643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,3072,0.744730684492323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,2560,0.6757946544223361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,2048,0.45198313395182294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,2048,0.553735097249349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,1536,0.35387465688917374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,2560,0.6337199740939671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,2048,0.5262906816270616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,1024,0.20996177196502686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,1536,0.43625601132710773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,1024,0.32664444711473256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,768,0.17221689224243164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,1536,0.43465243445502383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,512,0.13775822851392958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,1024,0.31801777415805393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,768,0.2582977877722846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,768,0.262113783094618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,256,0.09885778029759724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,512,0.21923289034101698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,128,0.08344088660346137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,512,0.21096977922651503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,256,0.19566667079925537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,256,0.1635768943362766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,64,0.07239555650287204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,51200,32,0.07635288768344455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,128,0.18369332949320474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,64,0.18711911307440865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,51200,128,0.12731999821133083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,51200,32,0.1923982169893053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,51200,3.560357411702474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,65536,4.448988596598308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,65536,6.581960042317708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,65536,8.414499070909288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,16384,1.113859600490994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,16384,1.8743404812282984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,51200,6.569950951470269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,51200,5.143828497992621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,12288,0.8357173071967231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,12288,1.327204492357042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,16384,1.67340268029107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,10240,0.6629075474209255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,12288,1.2462835311889648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,10240,1.0914790895250108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,10240,1.0396426518758137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,8192,0.8511857986450195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,8192,0.5552310943603516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,7168,0.49074400795830625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,8192,0.8406160142686633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,7168,0.7495111359490289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,6144,0.4334426720937093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,7168,0.7399768829345703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,5120,0.3558951218922933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,6144,0.6391990979512533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,6144,0.640775998433431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,5120,0.5429449081420898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,4096,0.28205156326293945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,5120,0.5417493184407552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,3584,0.23409777217441133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,4096,0.42171645164489746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,3072,0.20468800597720674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,4096,0.4347084363301595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,3584,0.3832844363318549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,3584,0.36729245715671116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,3072,0.31545334392123753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,2560,0.18494133154551187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,3072,0.333076450559828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,2048,0.14625333415137395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,2560,0.26875821749369305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,2560,0.28271910879347056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,1536,0.10746400223837958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,2048,0.248760011461046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,2048,0.22060622109307182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,1024,0.08504088719685872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,1536,0.1735502216551039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,1536,0.18759555286831328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,768,0.05973333120346069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,1024,0.14259555604722765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,512,0.049344887336095176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,1024,0.13044622209337023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,768,0.11851289537217881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,768,0.10374399688508774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,256,0.035134222772386335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,512,0.08589333295822144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,128,0.030863109562132094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,512,0.09586222304238214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,256,0.07440000110202365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,256,0.07351733578575982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,64,0.026179555389616225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,16384,32,0.02718044486310747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,128,0.06880178054173787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,16384,128,0.058865778976016574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,64,0.07007466422186957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,16384,32,0.07135822375615437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,51200,2.6554648081461587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,65536,3.4564569261338978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,65536,6.569427490234375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,65536,5.628682878282334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,16384,0.8767404556274414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,51200,5.289156595865886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,16384,1.5156923929850261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,51200,4.421491410997179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,12288,0.6632622083028158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,16384,1.4359440273708766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,10240,0.5304977628919814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,12288,1.084843529595269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,8192,0.4193262259165446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,12288,1.0739101833767362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,10240,0.8932035234239367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,10240,0.8969760470920138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,7168,0.3588346640268962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,8192,0.6999831199645996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,8192,0.7243990898132324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,6144,0.30377689997355145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,7168,0.6167439884609646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,5120,0.26343290011088055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,7168,0.6392781999376085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,6144,0.5267937978108724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,6144,0.5514142248365614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,4096,0.1934408876630995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,5120,0.45198843214246964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,5120,0.46675999959309894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,4096,0.34706844223870176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,3584,0.17724533875783285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,4096,0.3754675653245714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,3072,0.15271466308169895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,3584,0.32887199189927846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,3072,0.25846756829155815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,3584,0.3026977909935845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,2560,0.1357075505786472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,3072,0.2870568964216444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,2048,0.10850844118330215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,2560,0.21975110636817086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,1536,0.08081689145829943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,2048,0.18075199921925864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,2560,0.24229955673217773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,2048,0.20948178238338896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,1024,0.059225777784983315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,1536,0.1424657768673367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,1536,0.160936885409885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,1024,0.10660711261961196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,1024,0.12303466267055935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,768,0.08511555857128567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,768,0.10192977719836765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,512,0.03554133243030972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,512,0.07068710856967501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,256,0.027449778384632532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,512,0.08276977803972033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,256,0.060119999779595264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,128,0.022327999273935955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,256,0.06317155228720771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,768,0.04885422190030416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,64,0.0203155560625924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,12288,128,0.05040355523427328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,64,0.055642664432525635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,12288,32,0.021475555168257818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,32,0.05654488669501411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,12288,128,0.055721779664357506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,65536,2.8262587653266062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,51200,2.2971608903672958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,65536,5.926381429036458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,16384,0.7290613386366102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,65536,5.112435658772786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,51200,4.550240834554036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,12288,0.5445484585232204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,16384,1.335534201727973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,16384,1.3018639882405598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,12288,0.9612177742852105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,10240,0.4544915623135037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,51200,3.9945767720540366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,12288,0.9753946728176541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,8192,0.35195112228393555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,10240,0.792824904123942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,10240,0.8152977625528971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,8192,0.6234640015496148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,7168,0.30457690027025014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,8192,0.6576933330959743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,6144,0.26160089174906415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,7168,0.5487724410163032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,7168,0.5810089111328125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,6144,0.47121508916219074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,5120,0.20540266566806367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,4096,0.17365422513749865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,6144,0.5016080008612739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,5120,0.3922168943617079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,5120,0.4244524372948541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,3584,0.14729955461290148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,4096,0.3105226622687446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,3072,0.12663643889957005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,4096,0.3418960041469998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,3584,0.2688257694244385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,3584,0.299270232518514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,2560,0.1059546669324239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,3072,0.23122665617201063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,2048,0.0879511104689704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,3072,0.2603333261277941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,2560,0.1960995594660441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,2560,0.21952533721923828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,1536,0.0653182201915317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,2048,0.16089599662356907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,1536,0.12729599740770128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,1024,0.04799911048677233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,1536,0.14639555083380804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,768,0.04192177785767449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,1024,0.11132889323764378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,1024,0.09511644310421413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,512,0.0292231109407213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,768,0.0760257773929172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,768,0.09298400084177653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,256,0.02309777836004893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,512,0.07460266351699829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,512,0.06272000074386597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,128,0.018617777360810172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,256,0.057479110029008657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,128,0.048561778333452016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,64,0.017222222354676988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,128,0.045591112640168935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,10240,32,0.017507554756270517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,10240,2048,0.1843928893407186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,64,0.048707554737726845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,32,0.04933777782652113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,10240,256,0.052695112095938794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,65536,2.254293229844835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,51200,1.8680854373508031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,65536,5.0018187628851996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,65536,4.712760077582465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,51200,3.9420704311794705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,16384,1.1602986653645833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,16384,0.6035635736253526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,51200,3.6837493048773875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,12288,0.4386462105645074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,16384,1.19850402408176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,10240,0.3941368791792128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,12288,0.8336746427747938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,8192,0.3102693292829725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,12288,0.899263064066569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,10240,0.6923671298556857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,10240,0.7508711285061307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,8192,0.5464960204230415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,7168,0.2704533206091987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,8192,0.6082773208618164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,6144,0.23617243766784668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,7168,0.48176887300279403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,7168,0.5350906583997939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,6144,0.4135564433203803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,5120,0.18668622440761992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,6144,0.4740702311197917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,5120,0.34542134073045516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,5120,0.39088087611728245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,4096,0.15385156207614475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,3584,0.1384613381491767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,4096,0.271196444829305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,4096,0.31432443194919163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,3072,0.11831822660234238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,3584,0.2752844492594401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,3072,0.2024444474114312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,2560,0.09669511185752021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,3072,0.2378364404042562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,2560,0.17213955190446642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,2560,0.2017822265625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,2048,0.07862666580412123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,2048,0.14128355185190836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,2048,0.16821066538492838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,3584,0.23717066976759169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,1536,0.11137599415249294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,1024,0.04405422343148125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,1536,0.1343911091486613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,768,0.03508088986078898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,1024,0.08339289161894058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,1024,0.10275822215610081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,768,0.06691288948059082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,768,0.08516533507241143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,512,0.02460799945725335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,256,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,512,0.06847644514507718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,512,0.0550746652815077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,128,0.017217778497272067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,256,0.052298665046691895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,256,0.04526133338610331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,128,0.04044444362322489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,64,0.014968888627158271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,8192,128,0.04109777675734626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,32,0.015215999550289579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,64,0.03833422064781189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,8192,32,0.03913244605064392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,8192,1536,0.06357244650522868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,65536,2.101711908976237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,51200,1.760688887702094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,65536,4.624179416232638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,65536,4.445169660780165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,16384,0.5605848630269369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,51200,3.636028289794922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,12288,0.42113600836859805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,16384,1.070922639634874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,51200,3.4764097001817493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,16384,1.1298738055759006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,10240,0.3581404421064589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,12288,0.7761911286248101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,10240,0.6402408811781142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,12288,0.8492693371242948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,8192,0.2671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,7168,0.24248798688252768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,10240,0.7120249006483289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,8192,0.5091404385036892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,8192,0.5750586721632216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,6144,0.19460888703664145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,7168,0.47147464752197266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,7168,0.5075662400987413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,5120,0.16220355033874512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,6144,0.3851786719428168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,6144,0.43872621324327254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,4096,0.13677333460913763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,5120,0.3214942349327935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,3584,0.11674488915337457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,5120,0.3712426556481256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,4096,0.25299734539455837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,4096,0.2973751227060954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,3072,0.10011200110117595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,3584,0.21952088673909506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,2560,0.08453422122531468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,3072,0.18838400310940215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,3584,0.26028532452053493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,3072,0.22403200467427573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,2048,0.07329955365922716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,2560,0.160699552959866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,2560,0.19023199876149496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,1536,0.05447466505898369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,2048,0.13134222560458714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,1024,0.03810133205519782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,2048,0.1589448849360148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,1536,0.10392266511917114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,1536,0.1270951165093316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,768,0.030459556314680312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,1024,0.07775200075573392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,512,0.022637334134843614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,1024,0.0968853301472134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,768,0.0805893341700236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,256,0.018441778090265062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,512,0.05048622356520759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,512,0.0655768911043803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,256,0.039981332090165876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,128,0.015223999818166098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,256,0.0503386656443278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,64,0.019242667489581637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,128,0.035644445154401995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,7168,128,0.03945422172546387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,7168,32,0.015929778416951496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,64,0.034677333301968045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,32,0.03462400039037069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,7168,768,0.06257777743869357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,65536,1.9060516357421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,51200,1.3035164939032662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,65536,4.2247030470106335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,16384,0.43285332785712344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,65536,4.175555759006076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,51200,3.3224525451660156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,16384,0.9787173800998263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,12288,0.3331955538855659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,16384,1.0644764370388455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,10240,0.2996826701694065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,51200,3.2713529798719616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,12288,0.7147973378499349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,8192,0.22858932283189562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,12288,0.7987128893534342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,10240,0.6116586791144477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,7168,0.19786489009857178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,10240,0.669923570421007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,8192,0.4720444679260254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,8192,0.5411581993103027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,6144,0.16597867012023926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,7168,0.4148782094319661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,5120,0.13983733124203152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,7168,0.4781884617275662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,6144,0.3572639889187283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,4096,0.11150044865078396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,6144,0.4142835670047336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,5120,0.29763110478719074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,5120,0.3488062222798665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,3584,0.1001724468337165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,4096,0.23385511504279244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,3072,0.08520978026919895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,3584,0.20267200469970703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,4096,0.28072889645894367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,3584,0.2449644406636556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,2560,0.07547733518812391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,3072,0.1732462247212728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,3072,0.2111119959089491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,2048,0.06253955761591594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,2560,0.14850755532582602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,1536,0.047741333643595375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,2048,0.12170044581095378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,2560,0.17938044336107042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,2048,0.14902577135297987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,1024,0.034120000070995755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,1536,0.09627733627955119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,768,0.026510222090615168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,1536,0.11930667029486762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,1024,0.07166133324305217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,512,0.020236444142129686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,1024,0.09159288803736369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,768,0.07615200016233656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,768,0.057257778114742704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,256,0.017900443739361234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,512,0.04528266522619459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,128,0.01276711126168569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,256,0.03531733486387465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,256,0.0462666650613149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,64,0.014496889379289416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,128,0.03151466780238681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,6144,32,0.014779556128713818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,128,0.03600710961553786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,64,0.030910223722457886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,6144,32,0.03054666519165039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,6144,512,0.06126755475997925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,65536,1.617934226989746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,51200,1.269378662109375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,65536,3.9045935736762156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,16384,0.42253155178493923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,65536,3.9063254462348094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,51200,3.07204098171658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,12288,0.3041413360171848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,16384,0.8979750739203559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,16384,0.9959102206759982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,12288,0.651184876759847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,51200,3.060095045301649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,10240,0.24436354637145996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,8192,0.2092133363087972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,12288,0.7515599992540148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,10240,0.542101330227322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,7168,0.16719555854797363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,10240,0.6303004688686794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,8192,0.43834487597147626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,8192,0.50910308625963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,6144,0.13980178038279215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,7168,0.38150665495130753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,5120,0.11785156197018093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,7168,0.44994046952989364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,6144,0.3280293411678738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,6144,0.3883333206176758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,4096,0.09547822342978583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,5120,0.2741333378685845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,3584,0.08461333645714654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,4096,0.2146782212787204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,4096,0.2629786597357856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,5120,0.32898932033114964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,3072,0.07474755578570895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,3584,0.18629777431488037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,3584,0.22926934560139975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,2560,0.06344799862967597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,3072,0.1594408882988824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,3072,0.19774044884575737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,2560,0.13424800501929388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,2560,0.16815377606285942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,2048,0.11150666077931722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,1536,0.03906666570239597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,2048,0.14018489254845515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,1536,0.08827466434902614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,1024,0.02753866712252299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,1536,0.11241155200534397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,1024,0.06525599956512451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,768,0.02241244415442149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,2048,0.05344266692797343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,1024,0.0859804418351915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,768,0.05177422364552816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,768,0.07111999723646376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,512,0.017626666360431247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,256,0.017243555850452848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,512,0.05795822540918986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,256,0.04368977745374044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,256,0.031228442986806233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,128,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,5120,128,0.03296711047490438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,64,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,128,0.028580443726645574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,5120,32,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,64,0.027620444695154827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,32,0.027574222948816087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,5120,512,0.040623999304241605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,65536,1.1528213289048936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,51200,0.9151173697577583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,65536,3.477169672648112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,16384,0.3012506696912977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,51200,2.7038489447699656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,65536,3.769317203097873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,51200,2.9477361043294272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,16384,0.802284452650282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,12288,0.22060622109307182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,10240,0.20518933402167427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,16384,0.9622435039944119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,12288,0.5856915579901801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,8192,0.14996177620357937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,12288,0.7258017857869467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,10240,0.4904373486836751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,10240,0.6080631150139703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,7168,0.1395146714316474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,8192,0.3944239881303575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,6144,0.11173778110080296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,8192,0.4916790856255426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,7168,0.3475111060672336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,5120,0.09410844246546428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,7168,0.43452800644768613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,6144,0.2999520036909315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,4096,0.08020266559388903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,6144,0.3760053316752116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,5120,0.24973599116007486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,3584,0.07200977537367079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,4096,0.1967351171705458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,4096,0.25253332985772026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,3072,0.060270223352644176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,3584,0.1696444484922621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,3584,0.22016799449920654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,2560,0.053152889013290405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,3072,0.14495378070407444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,3072,0.18937244680192736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,2048,0.043290668063693576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,2560,0.12167644500732422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,2560,0.16138133737776014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,1536,0.03365244468053182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,2048,0.13378132714165583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,5120,0.3174195554521349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,1536,0.07961599694357978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,1024,0.023716444770495098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,1536,0.10754222340053982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,768,0.0188746667570538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,1024,0.05791822406980726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,1024,0.08211021953158908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,512,0.014845333165592618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,768,0.04639644424120585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,768,0.06820178031921387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,256,0.01092266705301073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,512,0.0533884432580736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,512,0.035735110441843666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,256,0.027494221925735474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,128,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,256,0.04030311107635498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,2048,0.09948977496888901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,64,0.0090524446633127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,4096,32,0.009011555876996782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,128,0.02474755545457204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,64,0.024121777878867254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,4096,128,0.03188622328970168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,4096,32,0.024337778488794964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,51200,0.8580800162421333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,65536,1.0869706471761067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,65536,3.258310317993164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,51200,2.4853831397162542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,16384,0.2800515492757161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,65536,3.634818606906467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,16384,0.7458657688564725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,12288,0.20547021759880912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,16384,0.9287502500745984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,12288,0.5566684405008951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,10240,0.17595377233293322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,12288,0.7016240225897895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,8192,0.14423288239373103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,10240,0.4660791291130914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,10240,0.5862248738606771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,8192,0.37514400482177734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,7168,0.12884533405303955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,51200,2.847533331976997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,8192,0.4751617643568251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,6144,0.10728089014689128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,7168,0.33118756612141925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,6144,0.28487200207180446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,5120,0.08971911006503636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,7168,0.41906044218275285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,6144,0.36280179023742676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,4096,0.07526399691899617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,5120,0.23823110262552896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,3584,0.06965066989262898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,5120,0.3062257766723633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,4096,0.24369332525465223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,3072,0.06153777572843763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,3584,0.16208088397979736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,3584,0.2130133310953776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,2560,0.05382755398750305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,3072,0.18337066968282065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,3072,0.13847022586398655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,2560,0.11677777767181396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,2048,0.04299822118547228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,2560,0.15563110510508218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,1536,0.027994665834638808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,2048,0.0949431127972073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,2048,0.1299991077846951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,4096,0.18681955337524414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,1536,0.07588711049821642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,1024,0.023726221587922838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,768,0.019810666640599568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,1536,0.10403377479977077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,1024,0.0542755557431115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,1024,0.0786000026596917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,512,0.016364443633291457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,768,0.06557066573037042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,768,0.04360711243417528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,512,0.03327911098798116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,256,0.013474666410022311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,512,0.052858667241202466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,128,0.009373333719041612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,256,0.03861600160598755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,256,0.026134222745895386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3584,128,0.031175110075208876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,32,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,32,0.022392888863881428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,65536,0.8521822293599447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3584,64,0.011005333728260465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,64,0.023021333747439917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,51200,0.7045315636528863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3584,128,0.023289778166347083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,65536,3.011550267537435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,16384,0.23568177223205566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,51200,2.3026470608181424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,65536,3.5009030236138234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,12288,0.16681688361697725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,16384,0.7090231047736274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,51200,2.7436818016899953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,10240,0.16255288653903538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,16384,0.8960062662760416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,12288,0.5295466846889919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,8192,0.11683111720614964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,12288,0.676604429880778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,10240,0.44974666171603733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,7168,0.10073422061072455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,8192,0.35691200362311465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,10240,0.5675990846421983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,8192,0.45924801296657985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,6144,0.09126489029990302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,7168,0.4040151172214084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,6144,0.271433777279324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,6144,0.3510444429185655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,5120,0.22624354892306855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,4096,0.05831199884414673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,5120,0.2959875530666775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,4096,0.17743821938832602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,7168,0.3144151104821099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,3584,0.056736886501312256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,3584,0.15337688393063015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,4096,0.23516533109876844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,5120,0.07434578074349298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,3584,0.20605688624911836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,3072,0.04513333241144816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,2560,0.03961688942379422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,3072,0.17614132828182646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,2560,0.10947733455234104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,2048,0.03163911236657037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,2560,0.15014044443766275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,3072,0.13119644588894314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,1536,0.02516533268822564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,2048,0.08984266387091742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,2048,0.12537156211005315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,1024,0.019000889526473153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,1536,0.09989511304431492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,1536,0.06916889217164782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,768,0.015469332536061605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,1024,0.051198220915264554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,1024,0.07563466495937772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,512,0.01205244412024816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,768,0.06341777907477485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,768,0.04107200105985006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,512,0.03085333440038893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,512,0.051352888345718384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,256,0.023936000135209825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,128,0.008143111235565608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,256,0.03837600019243028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,128,0.021714667479197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,3072,128,0.030413332912656996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,64,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,32,0.009338666995366415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,64,0.02103288968404134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,3072,32,0.02093244426780277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,3072,256,0.009036444127559662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,65536,0.7430906825595431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,51200,0.6089315414428711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,65536,2.7760240766737194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,16384,0.19917778174082437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,51200,2.1699661678738065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,16384,0.6660364468892416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,65536,3.3706480662027993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,12288,0.14709333578745523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,51200,2.637019475301107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,10240,0.15229866239759657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,16384,0.8625946574740939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,12288,0.5010284317864312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,12288,0.6509217686123318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,8192,0.11676711506313747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,10240,0.41902489132351345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,7168,0.10177155335744222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,10240,0.5469902356465658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,8192,0.33826398849487305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,6144,0.08975199858347575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,8192,0.44272621472676593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,7168,0.29773423406812877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,5120,0.07382844554053412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,7168,0.3905244403415256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,6144,0.2567182117038303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,4096,0.061462223529815674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,6144,0.3378230995602078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,5120,0.21411644087897408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,5120,0.28504623307122123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,3584,0.053414222266938954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,4096,0.16781155268351236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,3072,0.04488355583614773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,4096,0.22643287976582846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,3584,0.14498399363623724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,3584,0.19763733281029594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,2560,0.03549333413441976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,3072,0.12387555175357395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,2048,0.02765422231621212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,3072,0.17046311166551378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,2560,0.10366400082906087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,2560,0.14422844515906438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,1536,0.021254221598307293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,2048,0.0843777789009942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,1024,0.01566844516330295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,1536,0.0646702249844869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,2048,0.1203262276119656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,1536,0.09629155529869927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,768,0.012650666965378655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,1024,0.04760977625846863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,512,0.01035733355416192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,1024,0.07271466652552287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,768,0.03804888990190294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,256,0.007999111380841997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,768,0.059691554970211454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,512,0.028889778587553237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,512,0.04843199915356106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,128,0.00737955586777793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,256,0.022648000054889258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,64,0.00796622203456031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,256,0.036228444841172956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,128,0.020255999432669747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2560,128,0.02858755654758877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2560,32,0.008288000192907121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,64,0.01962577799956004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2560,32,0.01928622192806668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,65536,0.5592097706264919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,51200,0.4412231180402968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,51200,1.9930604298909504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,65536,2.53422482808431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,16384,0.148364437950982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,51200,2.5272418128119574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,16384,0.627961794535319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,65536,3.224632051255968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,12288,0.11271733707851833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,10240,0.09840889109505548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,16384,0.8254311349656848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,12288,0.47368976804945206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,10240,0.39644267823961044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,12288,0.6256782213846842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,8192,0.08169688781102498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,8192,0.32007379002041286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,10240,0.5265297889709473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,7168,0.06730133295059204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,8192,0.42585333188374835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,6144,0.06065066655476888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,7168,0.2829057905409071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,7168,0.3750568760765924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,6144,0.24300800429450142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,5120,0.05070133341683281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,5120,0.20258666409386528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,4096,0.04263644417126974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,5120,0.2735706700219048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,6144,0.3251813252766927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,3584,0.037457777394188776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,4096,0.1592293315463596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,4096,0.21712888611687553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,3072,0.03122577733463711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,3584,0.13837156030866835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,3584,0.18959377871619332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,2560,0.0269768883784612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,3072,0.1173422204123603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,3072,0.1641368865966797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,2560,0.09755377637015449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,2560,0.13853689034779867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,2048,0.02247200078434414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,1536,0.018559111489189994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,2048,0.08003466659122042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,2048,0.1156346665488349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,1024,0.013199999928474426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,1536,0.09281511439217462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,1536,0.060991108417510986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,1024,0.043700443373786084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,768,0.011334222224023608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,512,0.009165333376990424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,768,0.05651199817657471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,768,0.0358862214618259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,512,0.02757511039574941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,256,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,512,0.045554667711257935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,128,0.006275555739800136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,256,0.020980444219377305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,256,0.03498399919933743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,128,0.01865866614712609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,64,0.006021333237489064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,2048,32,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,128,0.02787999974356757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,64,0.018205333087179396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,2048,1024,0.0697386662165324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,2048,32,0.01830222209294637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,65536,0.4700640042622884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,51200,0.35354844729105633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,51200,1.853378719753689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,65536,2.3647759755452475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,16384,0.12438844309912787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,16384,0.5892364184061686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,12288,0.08303288618723552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,51200,2.427446153428819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,16384,0.7942719989352756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,65536,3.0956062740749783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,10240,0.08093600140677558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,12288,0.4449440108405219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,8192,0.06468088759316339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,10240,0.372561772664388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,12288,0.6025662422180176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,7168,0.05492800143029955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,10240,0.5062159962124294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,8192,0.30117956797281903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,6144,0.04660177893108792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,8192,0.4097564485337999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,7168,0.2655964427524143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,7168,0.3617360062069363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,5120,0.03990488913324144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,6144,0.22840621736314562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,4096,0.03271022107866075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,6144,0.3123813205295139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,5120,0.1907271146774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,3584,0.029491557015313044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,4096,0.14949156178368464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,5120,0.2631466653611925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,3072,0.025479111406538222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,4096,0.20848355028364393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,3584,0.12896888785892063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,3584,0.18176178137461343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,2560,0.02230666743384467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,3072,0.15590311421288386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,2048,0.01865600049495697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,3072,0.10999821954303318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,2560,0.09174755546781753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,2560,0.1333671145968967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,1536,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,2048,0.07472444242901273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,1024,0.011013333168294696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,2048,0.11095911264419556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,1536,0.059191114372677274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,1536,0.0887404415342543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,768,0.008756444685988957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,1024,0.0409733321931627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,1024,0.06778755452897814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,512,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,768,0.033075554503334895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,768,0.054077333874172635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,256,0.005992888990375731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,512,0.044695999887254506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,512,0.026124444272783067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,256,0.019331556227472093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,128,0.005648889061477449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,256,0.03400266501638625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,64,0.005013333426581489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,128,0.016887111796273124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1536,32,0.0053688887920644544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1536,128,0.027067555321587458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,64,0.01660711069901784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1536,32,0.016160888804329764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,65536,0.3134168783823649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,51200,0.25616444481743705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,65536,2.1920106675889754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,16384,0.0767493314213223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,51200,1.716409789191352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,16384,0.5559964179992676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,12288,0.06074044439527723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,65536,2.968366199069553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,16384,0.7620177798800998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,10240,0.053095112244288124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,12288,0.4203920099470351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,12288,0.5773457951015896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,8192,0.042028443680869214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,10240,0.3530435562133789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,10240,0.48537953694661456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,7168,0.03674844569630093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,8192,0.2847466733720568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,8192,0.39313332239786786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,51200,2.3262346055772567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,6144,0.031509333186679415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,7168,0.2506311204698351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,5120,0.026437333888477747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,7168,0.34686488575405544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,6144,0.21625778410169813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,5120,0.18055200576782227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,4096,0.023059555225902136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,6144,0.300543122821384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,3584,0.02020711037847731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,5120,0.25198756323920357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,4096,0.14236444897121853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,4096,0.20056711302863228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,3072,0.018082666728231642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,3584,0.1741093397140503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,2560,0.015879111157523263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,3584,0.1227875550587972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,3072,0.10335110955768162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,3072,0.14963466591305202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,2048,0.013257777525318993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,2560,0.08654489119847615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,2560,0.12757155630323622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,2048,0.0708284444279141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,2048,0.10591555303997463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,1024,0.008943110704421997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,1536,0.05485599901941088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,1024,0.03807377815246582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,768,0.007662222617202335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,1024,0.06233688857820299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,768,0.030212445391549006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,512,0.00656088896923595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,768,0.052074667480256825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,512,0.023950222465727065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,256,0.0052879999081293745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,1536,0.01129066695769628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,512,0.04228977693451775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,128,0.004590222405062782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,256,0.01779911087618934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,256,0.031530665026770696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,64,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,1536,0.08495377831988865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,1024,128,0.0251528885629442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,1024,32,0.004368888835112254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,128,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,32,0.014509333504570855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,65536,0.23434578047858345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,51200,0.1857137812508477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,1024,64,0.014815110299322339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,51200,1.6534115473429363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,16384,0.06472355789608426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,65536,2.112717310587565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,16384,0.5370346705118815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,12288,0.050323555866877236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,16384,0.7619928783840604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,65536,2.969726138644748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,10240,0.04240622123082479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,12288,0.40614133410983616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,51200,2.324127197265625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,8192,0.03257955445183648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,10240,0.34094667434692383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,12288,0.5781448682149252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,7168,0.028961777687072754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,10240,0.4850337770250108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,8192,0.27531199985080296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,8192,0.39301334487067324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,6144,0.025457777910762366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,7168,0.24270754390292695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,7168,0.34615998797946507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,5120,0.021799999806616042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,6144,0.20892266432444254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,5120,0.17462310526106092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,6144,0.3003164397345649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,4096,0.01796355512407091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,5120,0.253127998775906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,3584,0.01645599967903561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,4096,0.19999999470180937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,4096,0.13691911432478163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,3072,0.014517333772447375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,3584,0.11727555592854817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,3584,0.17431999577416313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,2560,0.012465777496496836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,3072,0.10001155402925278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,3072,0.14997243881225586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,2048,0.010999110837777456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,2560,0.08355822165807088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,2560,0.127402663230896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,2048,0.06808977656894259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,1536,0.009116444322797988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,2048,0.10589333375295003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,1024,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,1536,0.0517448882261912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,1536,0.08439822329415216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,1024,0.03603733248180813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,768,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,1024,0.06304977999793158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,512,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,768,0.02920000089539422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,768,0.05180266830656263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,256,0.004628444297446145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,512,0.023080888721677992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,512,0.04148533278041416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,128,0.004281777888536453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,256,0.03164888752831353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,256,0.017218665944205392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,128,0.014849777022997538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,64,0.003972444269392225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,768,128,0.024806222981876794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,768,32,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,64,0.014118222726715935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,768,32,0.013904000322024027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,65536,0.1718133290608724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,51200,0.1385164393319024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,51200,1.5930621888902452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,65536,2.0303217569986978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,16384,0.04755199948946635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,51200,2.2244088914659286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,16384,0.5170719888475206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,65536,2.837728076510959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,16384,0.7290266884697808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,10240,0.031038221385743883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,12288,0.39168887668185765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,12288,0.552095996008979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,10240,0.3292257785797119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,8192,0.02330044408639272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,10240,0.46427732043796116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,12288,0.03671822283003066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,8192,0.2662257883283827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,8192,0.3762142128414578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,6144,0.017685333887736004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,7168,0.23396534389919707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,6144,0.20137066311306426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,7168,0.3316817813449436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,5120,0.015358222855461968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,6144,0.2866773340437147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,4096,0.012843555874294706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,5120,0.24085601170857748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,5120,0.1685111125310262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,4096,0.13192799356248644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,3584,0.011707555916574268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,7168,0.020720889170964558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,3072,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,4096,0.1909359958436754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,3584,0.11306133535173203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,2560,0.009611555271678502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,3072,0.09584710995356242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,3584,0.1670844422446357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,2048,0.008045333127180735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,3072,0.1423706610997518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,2560,0.07976088921229045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,2560,0.12123910586039226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,1536,0.007181333170996771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,2048,0.06538044744067721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,1024,0.005632888939645555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,1536,0.08051200045479669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,1024,0.059077335728539355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,1024,0.033971554703182645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,768,0.005232000102599462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,768,0.04934222168392605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,768,0.02810844447877672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,512,0.004598222259018156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,512,0.03976266582806905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,256,0.0041804446114434134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,256,0.016197333733240765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,1536,0.04914044340451559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,256,0.030309332741631403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,128,0.0038337777886125776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,128,0.013853333062595792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,128,0.023686221904224817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,64,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,512,0.022287110487620037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,512,32,0.0038533334930737815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,64,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,512,2048,0.10054933362536961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,65536,0.133096006181505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,512,32,0.013253333667914072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,51200,0.10469866461224026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,65536,1.9538008371988933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,51200,1.5323030683729384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,16384,0.03826311230659485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,16384,0.5000613530476888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,65536,2.8265404171413846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,12288,0.02899200055334303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,16384,0.7253360218471951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,10240,0.02130488885773553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,12288,0.3781582249535455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,12288,0.5495893160502116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,8192,0.016826666063732572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,10240,0.31761421097649467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,10240,0.4620595508151584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,7168,0.015025777949227227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,8192,0.2561902205149333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,51200,2.2115741305881076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,8192,0.3741626739501953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,6144,0.013415999710559845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,7168,0.22528176837497285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,5120,0.011648889217111798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,7168,0.32987732357449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,6144,0.19467643896738687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,6144,0.2848942279815674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,4096,0.010064000056849586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,5120,0.16223822699652776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,5120,0.23951376809014213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,3584,0.00904444439543618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,4096,0.1275955571068658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,3072,0.00868088917599784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,3584,0.10953244235780503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,3584,0.16556889481014675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,4096,0.19089955753750273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,2560,0.007725333174069722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,3072,0.09162311421500312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,2048,0.006586666736337874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,3072,0.14127200179629856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,2560,0.0768995549943712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,2560,0.12002755535973443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,1536,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,2048,0.0622239973809984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,2048,0.09998310936821832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,1024,0.004932444542646408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,1536,0.04607199960284763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,768,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,1536,0.0780693358845181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,1024,0.03219822380277846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,1024,0.058295110861460366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,512,0.0041831110914548235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,768,0.026499556170569524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,768,0.0486968888176812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,256,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,512,0.021130666136741638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,512,0.03963911202218797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,128,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,256,0.029517332712809246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,64,0.0032506665835777917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,256,0.015522667103343539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,128,0.013072888884279462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,256,128,0.023339556323157415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,256,32,0.0033031110134389666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,65536,0.0965119997660319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,64,0.012503110700183444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,256,32,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,51200,0.07624977827072144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,51200,1.500565316942003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,16384,0.030213332838482324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,65536,1.9201484256320531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,16384,0.49010753631591797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,12288,0.021503110726674397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,51200,2.2134933471679688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,16384,0.7250737614101834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,65536,2.8233494228786893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,10240,0.015263999501864115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,12288,0.5496960216098362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,8192,0.012796444197495779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,10240,0.31079199579026967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,10240,0.46200980080498594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,8192,0.25084445211622447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,7168,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,8192,0.3731715679168701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,6144,0.010390222072601318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,7168,0.22092533111572266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,12288,0.370743989944458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,7168,0.32957156499226886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,5120,0.009302222066455418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,6144,0.1905439959632026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,4096,0.007620444728268518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,6144,0.28481867578294545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,5120,0.23945066663953993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,4096,0.12437955538431804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,5120,0.15939554903242323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,3584,0.007295110987292395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,4096,0.189940439330207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,3584,0.16450311077965632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,3072,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,3584,0.1074720025062561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,2560,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,3072,0.08996888664033677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,2560,0.07544710901048449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,2048,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,2560,0.1195288896560669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,2048,0.0989253322283427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,2048,0.06145955456627739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,1536,0.0052977779673205475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,1536,0.07909155554241605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,1024,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,1536,0.045520000987582736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,1024,0.031425777408811785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,3072,0.14106488227844238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,768,0.003863999827040566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,768,0.026391999589072332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,768,0.04801422357559204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,512,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,512,0.020576889316240948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,512,0.03878577881389194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,256,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,256,0.02895911203490363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,256,0.014818666709793938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,128,0.0031635556370019913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,128,0.012468444804350534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,128,0.022750222020679053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,64,0.003002666764789157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,128,32,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4352,128,1024,0.05760800176196628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,64,0.012075555821259817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,128,32,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,65536,0.09967288706037734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,51200,0.07435555590523614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,16384,0.02915289004643758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,12288,0.020474667350451153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,16384,0.4908284611172146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,10240,0.014851555228233337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,12288,0.370885345670912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,8192,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,10240,0.31076889567905003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,7168,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,8192,0.25129956669277614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,51200,1.5005982716878254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,7168,0.22091910574171278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,65536,1.918152067396376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,6144,0.19036622842152914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,4096,0.007376000285148621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,5120,0.15887378321753606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,3584,0.006919999917348226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,4096,0.12460977501339382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,3072,0.006256888724035687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,3584,0.10654399792353313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,2560,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,3072,0.08968622154659694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,2048,0.005231110999981563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,6144,0.010035555395815108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,2560,0.07505155271954007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,1536,0.004588444613748127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,1024,0.00397777764333619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,1536,0.04458399944835239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,5120,0.008707555631796518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,768,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,512,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,1024,0.03158933255407546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,768,0.025872000389628943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,256,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,512,0.02028888960679372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,128,0.002995555599530538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,64,0.002971555623743269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,256,0.014476444986131458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,64,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,128,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,65536,0.09577866395314534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,32,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,64,0.012056888805495368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,51200,0.0747928884294298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,16384,0.02736888825893402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,64,2048,0.06064266628689236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,12288,0.018350222044520907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,16384,0.4900844362046983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,10240,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,12288,0.3702462249332004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,8192,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,10240,0.3107173442840576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,7168,0.009392889009581672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,8192,0.2508515516916911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,6144,0.008656889200210571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,51200,1.5019689136081273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,5120,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,7168,0.22091821829477945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,6144,0.1902995506922404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,4096,0.01000533335738712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,3584,0.00905955582857132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,65536,1.9182391696506078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,4096,0.12465866406758626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,5120,0.15905866358015272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,3072,0.007962666451931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,3584,0.10732711023754543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,2560,0.006937777830494775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,2048,0.005218666460778978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,2560,0.07528355386522081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,3072,0.09045510821872288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,2048,0.0609440008799235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,1536,0.004569777597983678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,1024,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,768,0.0038648889296584656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,1024,0.03121866782506307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,1536,0.04559822214974297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,512,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,768,0.026130666335423786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,256,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,128,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,256,0.014848000473446317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,512,0.0206995556751887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,128,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,64,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4352,32,32,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,32,0.01149866647190518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4352,32,64,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,16384,4.210967169867621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,12288,3.1840061611599393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,16384,5.941023084852431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,10240,2.7867128584120007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,12288,4.153669145372179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,51200,13.372982449001737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,10240,3.3590977986653647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,8192,2.2748080359564886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,7168,1.9602319929334853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,8192,2.5966498057047525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,7168,2.2588816748725042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,6144,1.6719937854342992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,6144,1.9075565338134766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,5120,1.3699572881062825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,4096,1.158582263522678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,5120,1.5939938227335613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,4096,1.260199122958713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,3584,0.943968031141493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,51200,20.472269694010418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,3584,1.1003253724839952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,3072,0.9479084014892578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,2560,0.7576577928331164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,2560,0.8056089083353678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,3072,0.8127031326293945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,2048,0.5539369053310818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,2048,0.6568106545342339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,1536,0.4022328853607178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,1024,0.2939679887559679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,768,0.21519733799828422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,1536,0.5196222199334039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,512,0.16398221916622585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,768,0.31491290198432076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,1024,0.38947378264533145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,256,0.14322043789757624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,128,0.11054844326443142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,512,0.2608497672610813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,256,0.2323706679873996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,64,0.09639822112189399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,128,0.23045778274536133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,65536,32,0.10220711098776923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,32,0.2323999934726291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,65536,64,0.22582133611043295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,51200,10.439889695909288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,65536,13.38636440700955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,16384,3.464805391099718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,51200,16.280771891276043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,12288,2.5282933976915145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,16384,4.730912102593316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,10240,2.1397323608398438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,12288,3.315619574652778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,8192,1.6843795776367188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,10240,2.6829982333713107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,65536,21.30572509765625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,7168,1.4112346437242296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,8192,2.057560814751519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,6144,1.3182249069213867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,7168,1.8082425859239366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,5120,1.022832022772895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,4096,0.909322632683648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,6144,1.5300203959147136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,5120,1.275700463189019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,4096,1.0084720187717013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,3584,0.7986293368869357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,3072,0.6891217761569552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,3584,0.8817351129319934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,2560,0.581982241736518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,2048,0.4132382339901394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,3072,0.7602293226453992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,2560,0.6715760231018066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,1536,0.31455111503601074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,1024,0.21506578392452666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,2048,0.5758026440938314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,1536,0.41623110241360134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,768,0.179312891430325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,1024,0.3226284450954861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,512,0.12876089413960776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,768,0.24616800414191353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,256,0.09733688831329346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,512,0.20819821622636583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,128,0.08727111419041951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,256,0.18600532743665907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,64,0.07633333073721991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,128,0.17538844214545357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,51200,32,0.08112621969646878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,64,0.17938755618201363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,51200,32,0.18468356132507324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,51200,3.2737982008192272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,65536,4.1917805141872835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,16384,1.119110213385688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,12288,0.8113190862867566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,16384,1.793375015258789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,10240,0.6616889105902778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,12288,1.270965364244249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,8192,0.5380524529351128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,10240,1.0427413516574435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,65536,8.024186028374567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,8192,0.8285537825690376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,7168,0.4704302151997884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,6144,0.40560089217291934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,7168,0.7147137853834363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,51200,6.267912970648871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,5120,0.3454631169637044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,6144,0.6107866499159071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,4096,0.29651199446784127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,3584,0.24749334653218588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,5120,0.5091662406921387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,4096,0.40650399525960285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,3584,0.3674160109625922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,3072,0.2090160051981608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,2560,0.1761902173360189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,3072,0.3003635671403673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,2048,0.1434088945388794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,1536,0.1108035577668084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,2560,0.2551822132534451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,1024,0.079821334944831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,2048,0.2093057764901055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,768,0.06263999806510077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,1536,0.1654568910598755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,512,0.04787733488612705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,1024,0.1237608856625027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,256,0.03675555520587497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,768,0.09881777895821466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,128,0.030619555049472388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,64,0.026759111218982275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,256,0.07066311438878377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,512,0.08301511075761583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,128,0.06626666916741265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,16384,32,0.028079112370808918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,64,0.06686933173073663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,16384,32,0.06907555792066786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,51200,2.596714655558268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,65536,3.147127999199761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,16384,0.8193750911288791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,12288,0.602692445119222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,16384,1.4436364703708227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,10240,0.4976675775316026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,51200,4.951130761040582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,8192,0.4108026557498508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,10240,0.8477617899576823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,65536,6.461684332953559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,7168,0.3575066725413005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,8192,0.6660844484965006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,12288,1.027091556125217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,5120,0.24670844607883027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,6144,0.30877778265211314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,4096,0.19495288530985513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,7168,0.5859208636813694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,6144,0.5032729042900933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,5120,0.41813156339857316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,4096,0.32769778039720326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,3072,0.1602026621500651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,3584,0.28756266170077854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,2560,0.13570311334398058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,2048,0.10657688644197251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,3072,0.24676889843410918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,1536,0.07733155621422662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,2560,0.21034577157762316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,2048,0.17246666219499376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,1024,0.05718666977352566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,768,0.044729779163996376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,1536,0.13578489091661242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,512,0.034613334470325045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,1024,0.10201333628760444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,3584,0.175599111451043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,256,0.02661688956949446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,768,0.08213422033521864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,128,0.02383022175894843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,512,0.06776266627841525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,64,0.020911999874644812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,12288,32,0.021710221966107685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,256,0.0572355588277181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,128,0.05336266756057739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,64,0.05346666773160299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,12288,32,0.054577777783075966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,51200,2.0983990563286676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,65536,2.63410398695204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,16384,0.683784008026123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,12288,0.4972693125406901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,16384,1.2698613272772896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,10240,0.4168257713317871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,12288,0.9111333423190646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,65536,5.495847066243489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,8192,0.3318195607927111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,7168,0.28913243611653644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,10240,0.7898364596896701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,6144,0.2531884511311849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,8192,0.5945013364156088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,51200,4.3462677001953125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,5120,0.2100728882683648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,4096,0.18702222241295707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,6144,0.4545484648810492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,3584,0.15109422471788195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,5120,0.37307198842366535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,3072,0.12735022438897028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,4096,0.2938008838229709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,3584,0.2547075483534071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,2560,0.10799200005001491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,2048,0.08494044674767388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,3072,0.21977244483100045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,1536,0.06667821937137179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,2560,0.18677511480119494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,7168,0.5226400163438585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,1024,0.04790400134192573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,768,0.04225422276390923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,1536,0.12100266085730658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,2048,0.15360800425211588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,512,0.028948442803488836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,1024,0.09085600243674384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,256,0.02273955610063341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,768,0.07294399870766534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,128,0.020255110330051847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,64,0.017871111631393433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,256,0.050591998630099826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,10240,32,0.01833866702185737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,128,0.046591109699673124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,64,0.046795556942621865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,32,0.047572443882624306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,10240,512,0.06052088737487793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,51200,1.679752031962077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,65536,2.190558327568902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,16384,0.5371866756015354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,12288,0.40689510769314235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,16384,1.0930124918619792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,10240,0.3361697726779514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,12288,0.7920977804395887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,8192,0.2801004515753852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,10240,0.6563244395785862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,7168,0.239757325914171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,8192,0.5215804311964247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,65536,4.736630333794488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,6144,0.2071839968363444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,5120,0.1751128832499186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,7168,0.4582471317715115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,6144,0.39391199747721356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,4096,0.14310399691263834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,3584,0.12667644023895264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,5120,0.32801156573825413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,4096,0.2565777831607395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,3072,0.113301330142551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,2560,0.08863555722766453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,3584,0.22408000628153482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,3072,0.1929831107457479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,2048,0.073216888639662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,1536,0.06032088730070326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,51200,3.7690230475531687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,2560,0.1635848946041531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,1024,0.04219466778967115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,2048,0.13446222411261663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,768,0.03375555409325494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,1536,0.1063377790980869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,512,0.0258240004380544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,1024,0.07958577738867866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,256,0.01958311100800832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,128,0.01702844434314304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,768,0.06444178024927776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,512,0.052611556318071157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,256,0.04353688822852241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,32,0.01514844430817498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,128,0.03825777769088745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,64,0.037485334608289934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,8192,32,0.036975112226274275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,8192,64,0.014771555860837301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,51200,1.4509732988145616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,65536,1.8223465813530817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,16384,0.4872346454196506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,12288,0.3569635550181071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,16384,1.0143768522474501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,10240,0.2979671160380046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,12288,0.7357218000623914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,8192,0.27325865957472056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,10240,0.6075821982489692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,8192,0.483791987101237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,7168,0.2266871134440104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,51200,3.4726151360405813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,6144,0.19089599450429282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,7168,0.4262684451209174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,5120,0.15619822343190512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,6144,0.36675111452738446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,4096,0.1248133314980401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,5120,0.30485600895351833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,3584,0.11053866810268825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,4096,0.23895910051133898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,65536,4.519502427842882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,3072,0.09527911080254449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,3584,0.20777689086066353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,2560,0.08021777868270874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,3072,0.178911116388109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,2048,0.07002222537994385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,1536,0.05322666631804573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,2560,0.15235822730594212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,1024,0.03579644362131754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,2048,0.1251128911972046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,768,0.029127111037572224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,1536,0.0997004442744785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,1024,0.07417688767115276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,512,0.02256977723704444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,768,0.059303998947143555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,256,0.017868444323539734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,128,0.014946666028764514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,512,0.04842222068044874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,64,0.012409778104888068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,256,0.03798666596412659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,7168,32,0.014204444156752693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,64,0.03345511025852627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,128,0.034987555609809026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,7168,32,0.03267466690805223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,51200,1.2325377994113498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,65536,1.5547786288791234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,16384,0.4136053456200494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,12288,0.32150223520067006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,16384,0.9347111384073893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,10240,0.2593173450893826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,12288,0.6856035656399198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,8192,0.20695111486646864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,10240,0.5615350935194228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,8192,0.4477333492702908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,7168,0.1804515520731608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,6144,0.16092177232106528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,65536,4.016599867078993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,51200,3.1681253645155163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,7168,0.3937226666344537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,5120,0.13426844278971353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,4096,0.11288266711764866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,6144,0.33931554688347715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,3584,0.09603999720679389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,5120,0.28279821077982586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,3072,0.0846337808503045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,2560,0.0683715542157491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,4096,0.22132266892327201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,3584,0.1960995594660441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,2048,0.054698665936787925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,3072,0.16426311598883733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,1536,0.042028443680869214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,1024,0.03045155604680379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,2560,0.14067467053731283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,2048,0.11599377791086833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,768,0.023939554889996845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,512,0.0188044442070855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,1536,0.09136888715955947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,256,0.014501333236694336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,768,0.05484088924196032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,128,0.01585155559910668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,512,0.04326577650176155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,64,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,256,0.03365155392222934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,6144,32,0.014474666780895658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,128,0.030538668235143025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,64,0.030188444587919448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,32,0.02979733215437995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,6144,1024,0.06844177511003283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,51200,1.0971626705593533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,65536,1.3557155397203233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,16384,0.37435823016696507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,12288,0.2779991096920437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,16384,0.8469760153028699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,10240,0.2333608998192681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,12288,0.6163226763407389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,8192,0.18560443984137642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,10240,0.514692465464274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,7168,0.1637599998050266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,8192,0.411448876063029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,6144,0.13820800516340467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,65536,3.7198274400499134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,7168,0.3620888921949599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,51200,2.872816933525933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,6144,0.3121359878116184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,4096,0.09376266929838394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,5120,0.25968533092074925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,3584,0.08386044369803534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,3072,0.07243288887871636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,4096,0.2034764422310723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,2560,0.06280444728003608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,3584,0.1758355564541287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,2048,0.05123200019200643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,2560,0.12795200612809923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,3072,0.1511733399497138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,1536,0.039742221434911094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,2048,0.10635644197463989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,5120,0.11663288540310329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,1024,0.02609511050913069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,768,0.0199946661790212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,1536,0.0845537781715393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,512,0.015881778465376962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,1024,0.06178577740987142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,768,0.04948888884650337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,256,0.01290933373901579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,128,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,512,0.03872355487611558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,256,0.03018666638268365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,64,0.009720889230569204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,128,0.02719466719362471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,5120,32,0.010766222245163388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,64,0.02712000078625149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,5120,32,0.02645955483118693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,51200,0.8347937795850965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,65536,1.0307653215196397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,16384,0.2799217700958252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,16384,0.7703635427686902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,12288,0.2058560053507487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,10240,0.1695999966727363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,12288,0.5561902258131239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,8192,0.13776889112260607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,10240,0.4649670918782552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,65536,3.2265928056504993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,7168,0.12202133072747125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,6144,0.10551110903422038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,51200,2.5861057705349393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,8192,0.3746391137440999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,5120,0.08962044450971816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,7168,0.33048889372083873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,4096,0.0736266639497545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,3584,0.06765955686569214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,6144,0.28416445520189076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,4096,0.18530221780141196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,5120,0.23666755358378092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,3072,0.058782219886779785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,2560,0.05029155479537117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,3584,0.16008622116512722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,2048,0.03992977738380432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,1536,0.031034668286641438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,3072,0.1374924447801378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,2560,0.11571555667453343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,1024,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,768,0.017272889614105225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,2048,0.09515644444359674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,512,0.01330488920211792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,1536,0.07663822174072266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,1024,0.05549955699178907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,256,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,768,0.04460266563627455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,512,0.03422844409942627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,64,0.008303999900817871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,256,0.026495999760097925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,128,0.024363555841975745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,32,0.00904799997806549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,64,0.023613333702087402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,4096,32,0.02364088926050398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,65536,0.9121422237820096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,51200,0.726542207929823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,4096,128,0.009387555221716562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,16384,0.2577279938591851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,12288,0.18822133541107178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,16384,0.7078071170383029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,12288,0.5274115668402778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,10240,0.4424995581309001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,8192,0.11914932727813721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,65536,3.1075333489312063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,8192,0.356154653761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,10240,0.1541057825088501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,51200,2.413505766126845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,7168,0.10875466797086929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,6144,0.09834400150511001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,5120,0.08170221911536323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,4096,0.06462488571802776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,7168,0.31419287787543404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,3584,0.05938133266237047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,6144,0.26999378204345703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,5120,0.22503911124335396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,3072,0.05185688866509331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,4096,0.17590399583180746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,2560,0.0429004430770874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,2048,0.03638133406639099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,3584,0.15296710862053767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,1536,0.02913155489497715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,2560,0.11006933450698853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,1024,0.019448000523779128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,3072,0.13175200091467962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,2048,0.09017511208852132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,768,0.0159582214223014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,1536,0.07245688968234591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,512,0.01293333371480306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,256,0.010004444254769219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,1024,0.05137333273887634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,768,0.041767110427220665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,128,0.00851733320289188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,512,0.032511999209721885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,64,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,256,0.02533333334657881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,128,0.022416889667510986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3584,32,0.008004444340864817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,64,0.022312000393867493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3584,32,0.022038223014937505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,65536,0.8227821985880533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,51200,0.6654746797349718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,16384,0.22784090042114258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,16384,0.6761751174926758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,12288,0.16825332906511095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,12288,0.5019155608283149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,65536,2.7657992045084634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,10240,0.13843288686540392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,8192,0.11298933294084336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,51200,2.1737982432047525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,7168,0.10380978054470485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,6144,0.08711910910076565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,10240,0.4202168782552083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,8192,0.33836089240180117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,5120,0.07218488719728258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,4096,0.0563653310139974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,7168,0.29861243565877277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,6144,0.25646665361192494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,3584,0.05034133460786608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,3072,0.04433955417739021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,5120,0.21386044555240205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,4096,0.1672346724404229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,2560,0.04014488723542955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,3584,0.14433777332305908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,2048,0.03296533226966858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,1536,0.025402666793929204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,3072,0.12445422013600667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,2560,0.1044808891084459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,1024,0.0181040002240075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,2048,0.0851831105020311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,768,0.014749333262443542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,512,0.011167999770906238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,1024,0.04853422111935086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,256,0.009015999734401703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,768,0.0392542216512892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,128,0.007977777885066139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,256,0.023369777533743117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,512,0.03019733230272929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,64,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,128,0.02095288866096073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,3072,32,0.008681777450773451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,64,0.021057777934604224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,32,0.020346666375796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,65536,0.7048533227708605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,51200,0.5580853356255425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,16384,0.18813067012363008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,16384,0.6313920021057129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,12288,0.13814132743411595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,3072,1536,0.06545244322882758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,10240,0.1100426647398207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,12288,0.4740791320800781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,8192,0.0957697762383355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,10240,0.3979208999209934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,65536,2.5807039472791886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,8192,0.32020799318949383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,6144,0.0704782207806905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,7168,0.28292444017198354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,5120,0.05502133236991035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,6144,0.2427911228603787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,4096,0.045174221197764076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,5120,0.20251911216311982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,4096,0.1579493284225464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,3584,0.04011022382312351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,3072,0.036072889963785805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,7168,0.07639911439683703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,2560,0.03068888849682278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,3584,0.13699021604326037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,2048,0.024149333437283833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,3072,0.11794932683308919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,2560,0.09886044263839722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,1536,0.02033955521053738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,1024,0.014207111464606391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,2048,0.08075200186835395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,768,0.011968000067604912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,1536,0.06167822413974338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,512,0.00960799968904919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,1024,0.04586933387650383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,256,0.007336888876226213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,768,0.03667733404371474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,512,0.028559111886554297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,128,0.007253333098358578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,64,0.006609777609507243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,256,0.022302221920755174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,128,0.019632889164818656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,64,0.018988443745507132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2560,32,0.006576889091067844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,51200,2.025180392795139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2560,32,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,51200,0.4270666705237494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,65536,0.5288195610046387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,16384,0.1433146662182278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,12288,0.11036622524261475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,16384,0.6012489000956217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,12288,0.44843022028605145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,10240,0.08877955542670356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,51200,1.8737475077311199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,8192,0.07577421930101183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,10240,0.3751804563734267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,65536,2.378845426771376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,7168,0.07213866710662842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,6144,0.06006044149398804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,8192,0.30269956588745117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,5120,0.051354666550954185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,7168,0.2668942345513238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,6144,0.22907733917236328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,4096,0.03988000088267856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,5120,0.19065599971347383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,3584,0.03509511219130622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,3072,0.030576000610987347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,4096,0.14872088697221544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,2560,0.026687999566396076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,3584,0.12872800562116835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,2048,0.022054221895005967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,3072,0.11056444380018447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,2560,0.09279555744594997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,1536,0.01700355609258016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,1024,0.0125591109196345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,1536,0.05814311239454481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,768,0.010441777606805166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,1024,0.040808889600965716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,512,0.008760000268618265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,768,0.034367998441060386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,256,0.007028444773621029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,512,0.026886221435334947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,128,0.006488889041874144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,256,0.020625778370433383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,64,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,2048,32,0.005984000033802456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,128,0.018616000811258953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,64,0.017613333132531907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,32,0.017588444881969027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,2048,2048,0.07603110869725545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,65536,0.4161662260691325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,51200,0.3399119906955295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,16384,0.10984800259272258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,12288,0.08565688795513576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,16384,0.5592000219557021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,10240,0.07341777616076998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,12288,0.4227200084262424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,8192,0.05905866622924805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,10240,0.3541777663760715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,7168,0.05180444320042928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,8192,0.2865653302934435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,51200,1.7367910808987086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,6144,0.04426222377353244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,5120,0.0373964442147149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,7168,0.25200178888109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,65536,2.2161733839246964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,4096,0.030969778696695965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,3584,0.026896889011065166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,6144,0.2169804440604316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,3072,0.0248364441924625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,4096,0.1410177813635932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,5120,0.18027999666002062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,3584,0.12226311365763347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,2560,0.021213332811991375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,2048,0.017700443665186565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,3072,0.10374399688508774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,1536,0.013756444056828817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,1024,0.010690666735172272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,2048,0.0711617800924513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,2560,0.08838844299316406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,1536,0.056102222866482206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,512,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,768,0.03233688738611009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,512,0.024951110283533733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,256,0.006685333533419504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,128,0.00526577772365676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,256,0.01887911061445872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,64,0.004950222041871813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,128,0.01684533390733931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,32,0.004941333499219683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,64,0.01620533400111728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,32,0.016209777858522203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1536,768,0.00868000007337994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,65536,0.3213004536098904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1536,1024,0.03870222303602431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,16384,0.08454310894012451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,51200,0.25542489687601727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,12288,0.06599289178848267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,16384,0.5215484301249186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,12288,0.3945546680026584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,10240,0.055996444490220815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,8192,0.044264889425701566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,10240,0.3307110998365614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,7168,0.037873778078291155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,8192,0.26758310529920787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,6144,0.03175910976197984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,51200,1.6080559624565973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,5120,0.026146666871176824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,7168,0.23508532842000326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,4096,0.021228444245126512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,6144,0.20265511671702066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,65536,2.0539964040120444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,3584,0.018935110833909776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,5120,0.16840622160169813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,3072,0.01665866706106398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,4096,0.1322737799750434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,2560,0.01462577780087789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,3584,0.11448533005184597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,3072,0.09657955831951565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,2048,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,2560,0.08138933446672228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,1536,0.010201777848932479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,1024,0.008029333419269985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,2048,0.06688177585601807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,768,0.007263999846246507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,1536,0.05099111133151584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,512,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,1024,0.03535644544495477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,768,0.02924799919128418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,512,0.023366222778956097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,128,0.004618666652176115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,256,0.017616000440385606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,64,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,128,0.015527110960748462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,32,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,64,0.014890667464998035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,1024,32,0.014825777875052558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,65536,0.2193200058407254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,51200,0.17590133349100748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,1024,256,0.005241777747869492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,16384,0.05906222263971964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,16384,0.5098071098327637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,12288,0.04520888792143928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,10240,0.03864977757136027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,51200,1.5634560055202906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,65536,2.000429365370009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,8192,0.03241777751180861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,7168,0.029140445921156142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,12288,0.3858782185448541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,6144,0.024777778320842322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,10240,0.3231911129421658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,5120,0.02124533388349745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,8192,0.2618186738755968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,7168,0.22990756564670137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,4096,0.017224889662530687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,3584,0.015268445014953613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,5120,0.16444533401065403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,3072,0.013637333280510373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,4096,0.1286897791756524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,6144,0.1982426643371582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,2560,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,3584,0.11100711425145467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,2048,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,1536,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,3072,0.09399644533793132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,1024,0.007263999846246507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,2560,0.0796826680501302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,1536,0.048885332213507764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,768,0.006312888943486744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,1024,0.03426133261786567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,512,0.005613333235184352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,256,0.004927999857399198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,768,0.028360890017615423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,128,0.004348444441954295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,512,0.02238044473859999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,64,0.004184000194072723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,256,0.016878222425778706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,128,0.014518222875065274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,768,32,0.004442666553788715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,64,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,32,0.013756444056828817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,65536,0.16679910818735758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,51200,0.13191378116607666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,16384,0.04633955491913689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,768,2048,0.06491822004318237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,12288,0.03523022267553542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,16384,0.49080446031358504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,12288,0.37087199423048234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,10240,0.029353777567545574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,8192,0.021690666675567627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,10240,0.3115484449598524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,8192,0.2523146735297309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,7168,0.018425777554512024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,6144,0.01551555593808492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,51200,1.5051822662353516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,7168,0.2216986682679918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,5120,0.013764444324705334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,65536,1.919192843967014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,6144,0.19061688582102457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,4096,0.011508444117175208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,3584,0.01033688916100396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,3072,0.009319110876984065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,5120,0.15855467319488525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,2560,0.008628444539176093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,4096,0.12404533227284749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,2048,0.007272000114123027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,3072,0.090302222304874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,3584,0.10744000143475002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,2560,0.07607022258970472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,1024,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,2048,0.06201599703894722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,1536,0.047331555022133716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,768,0.004952888935804367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,768,0.027403554982609216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,512,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,1536,0.006642666541867786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,512,0.021705778108702764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,256,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,128,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,128,0.013783999615245394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,256,0.016220443778567843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,64,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,1024,0.03251377741495768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,512,32,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,65536,0.11028622256384955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,32,0.012777778009573618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,512,64,0.013756444056828817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,51200,0.08650400241216023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,16384,0.037257777320014104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,16384,0.47193066279093426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,12288,0.0248391115003162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,12288,0.3576791021558974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,10240,0.02124711043304867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,51200,1.4457635879516602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,8192,0.015631111131774057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,65536,1.8453040652804906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,10240,0.29992977778116864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,7168,0.013304000099500021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,6144,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,5120,0.01036533299419615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,8192,0.24274844593471953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,6144,0.18393066194322374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,7168,0.2133164405822754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,4096,0.00906133320596483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,5120,0.15257867177327475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,3584,0.008465777668688033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,3072,0.007662222617202335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,2560,0.006923555499977536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,4096,0.11916800340016682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,2048,0.0063182223174307085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,3072,0.086535996860928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,2560,0.07270222240024142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,3584,0.10317689180374146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,1536,0.005671999934646819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,1024,0.004850666556093427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,768,0.004531555705600315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,1536,0.044248888889948525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,512,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,1024,0.031016889545652602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,256,0.0036115555299652945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,512,0.02071555621094174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,768,0.026537777649031744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,128,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,64,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,64,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,256,32,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,32,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,65536,0.09337333175871108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,2048,0.059049778514438205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,51200,0.0761537750562032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,256,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,16384,0.028836445675955877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,256,128,0.013219555219014486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,12288,0.020631111330456205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,16384,0.4630266825358073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,10240,0.015252444479200574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,12288,0.3503448963165283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,51200,1.4158435397677953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,8192,0.012768000364303589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,10240,0.2936791049109565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,7168,0.01146044416560067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,65536,1.809302224053277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,6144,0.010398222340477837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,5120,0.00908088849650489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,8192,0.23753778139750162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,7168,0.20892711480458578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,4096,0.007941332956155142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,6144,0.17981955740186903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,3584,0.007267555428875818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,3072,0.006587555425034628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,4096,0.11675911479526097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,5120,0.15015021959940592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,3584,0.09995200236638387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,2560,0.006287999865081575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,2048,0.00555288874440723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,3072,0.08445066875881618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,1536,0.00488711107108328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,1024,0.004210666649871402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,2048,0.057504885726504855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,1536,0.04310311211480034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,768,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,2560,0.07197244299782647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,1024,0.03049066662788391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,256,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,768,0.025788444611761305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,128,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,256,0.01480977733929952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,64,0.0031679999083280563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,512,0.020336889558368258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,128,0.01240355521440506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,32,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,64,0.01205244412024816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,65536,0.0895964437060886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,128,32,0.012075555821259817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,51200,0.07115999857584636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,128,512,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,16384,0.028373334142896865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,16384,0.46274402406480575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,12288,0.01678222252262963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,51200,1.4165777630276148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,65536,1.8085208468967016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,10240,0.013218666944238873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,8192,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,10240,0.29346932305230033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,7168,0.010825778047243753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,8192,0.2377635637919108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,6144,0.009787555370065901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,7168,0.208259556028578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,5120,0.008641777767075432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,12288,0.3499804337819417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,4096,0.007629333270920648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,6144,0.18020799424913195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,5120,0.14901333385043675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,3584,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,3072,0.006611555400821898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,4096,0.11640266577402751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,2560,0.005896889087226655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,3584,0.09972711404164632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,2048,0.0052764444715446895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,1536,0.004575110971927643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,3072,0.08515733480453491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,1024,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,2048,0.05736088752746582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,1536,0.04317155480384827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,768,0.003522666792074839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,512,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,768,0.025388444463411968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,256,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,1024,0.030255999830034044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,512,0.02020355562369029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,128,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,64,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,256,0.014471999473041959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,128,0.012404444317022959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,64,32,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,65536,0.08579288588629828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,32,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,64,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,51200,0.0691244469748603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,16384,0.026587555805842083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,64,2560,0.07095288568072848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,12288,0.013627555635240344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,16384,0.462083551618788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,12288,0.3495768970913357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,10240,0.011363555159833698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,8192,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,10240,0.2935306760999891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,8192,0.23691111140780977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,7168,0.010013333625263637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,6144,0.009006222089131674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,51200,1.4161110983954535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,7168,0.20843022399478486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,5120,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,6144,0.1796097755432129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,65536,1.808679156833225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,4096,0.011137777732478248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,3584,0.009017777939637503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,3072,0.008610666626029545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,5120,0.14963644080691868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,4096,0.11637600262959798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,2560,0.007261333366235097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,2048,0.005205333232879639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,3072,0.08507021930482651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,3584,0.1013608906004164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,1536,0.004629333400064045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,2560,0.07118222448560926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,2048,0.05753066804673937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,768,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,1024,0.029994666576385498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,1536,0.043134222428003945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,512,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,768,0.025368889172871906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,256,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,128,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,256,0.014482667048772177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,512,0.02030933399995168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,32,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,64,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,32,0.011789333489206104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4097,32,1024,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4097,32,128,0.012073777616024017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,16384,4.127605438232422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,16384,5.876589457194011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,51200,12.912634107801649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,12288,3.1526319715711804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,16384,4.446481068929036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,51200,15.112233479817709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,51200,20.022310045030384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,12288,4.087870279947917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,12288,3.2190382215711804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,10240,2.6673732333713107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,10240,3.327040990193685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,8192,2.1232115427652993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,10240,2.698350270589193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,8192,2.53092532687717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,8192,2.1722284952799478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,7168,1.7547618018256292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,7168,2.2212132347954645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,6144,1.6462355719672308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,6144,1.8773920271131728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,7168,1.9057013193766277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,5120,1.2706338034735787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,6144,1.6293964385986328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,5120,1.5704675250583227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,4096,0.9933048884073893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,5120,1.3771582709418402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,4096,1.2409946653578017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,3584,0.9222408930460612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,4096,1.1162595748901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,3584,1.0914071400960286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,3072,0.8130222426520454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,3584,0.9853892856174045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,3072,0.9360230763753256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,2560,0.6276417838202583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,3072,0.8625849088033041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,2048,0.5411421987745497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,2560,0.7971875402662488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,2560,0.7317591243320041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,1536,0.36206666628519696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,2048,0.6520115534464518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,2048,0.608566231197781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,1536,0.5127111011081272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,1024,0.271757337782118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,1536,0.4854355388217502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,768,0.22022667196061876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,1024,0.38609332508511013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,1024,0.3676915433671739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,768,0.30386755201551646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,512,0.16758310794830322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,256,0.12150488959418403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,512,0.25788977411058217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,512,0.24331288867526582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,128,0.10115289025836521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,256,0.2304417822096083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,256,0.19021777311960855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,64,0.08432000213199192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,128,0.14727111657460532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,65536,32,0.0866853329870436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,64,0.22210221820407441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,32,0.22950755225287545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,65536,768,0.3045182228088379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,65536,128,0.2179644505182902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,51200,10.099763658311632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,65536,13.001146952311197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,65536,20.024234347873264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,65536,15.337378607855904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,16384,3.2811145782470703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,16384,4.700638241238065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,51200,16.321204291449654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,12288,2.4327235751681857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,51200,11.80712890625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,16384,3.593382941351997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,12288,3.236336814032661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,10240,2.174104902479384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,10240,2.6449546813964844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,12288,2.6493074629041886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,8192,1.6251786549886067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,8192,2.0175181494818792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,10240,2.189769744873047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,7168,1.4489146338568792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,8192,1.7664035161336262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,7168,1.7745466232299805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,6144,1.27437867058648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,7168,1.5381476084391277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,6144,1.5140115949842665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,5120,0.9937982559204102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,4096,0.7673377990722656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,5120,1.2596764034695096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,6144,1.3300409317016602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,5120,1.1252880096435547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,3584,0.7184284528096517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,4096,0.9938302569919162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,3584,0.8712213304307727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,4096,0.9108497831556531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,3072,0.5972595744662815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,3584,0.8041795624627007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,2560,0.5267857975429958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,3072,0.7527297867668999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,3072,0.7026791042751737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,2560,0.6376150978936089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,2048,0.388915565278795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,1536,0.29450488090515137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,2560,0.6038444307115343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,2048,0.5221901999579536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,2048,0.49628978305392796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,1024,0.21621421972910562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,1536,0.41055644883049863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,1536,0.39540889528062606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,1024,0.3087235556708442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,768,0.15335199568006727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,1024,0.30057067341274685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,512,0.12231822808583577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,768,0.24410933918423125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,768,0.24888534016079375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,512,0.20663377973768446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,256,0.09434755643208821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,256,0.18395466274685332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,256,0.1565271086162991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,128,0.17389067014058432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,512,0.19890666007995605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,64,0.07602044608857897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,51200,128,0.12074222829606797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,32,0.08102311028374566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,51200,128,0.08141244120068021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,64,0.17678756184048125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,51200,32,0.1827546623018053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,65536,3.9295760260687937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,51200,3.206253263685438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,65536,7.940917121039496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,65536,6.154158274332683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,16384,1.0542710622151692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,16384,1.7655501895480687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,51200,6.171668582492405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,51200,4.804087320963542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,12288,0.7383511331346301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,16384,1.5671973758273656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,12288,1.2469768524169922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,10240,0.611628426445855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,8192,0.4976515769958496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,12288,1.1604497697618272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,10240,1.0266062418619792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,7168,0.45750845803154844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,10240,0.9746969011094836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,8192,0.8020444446139865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,8192,0.7868213123745389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,6144,0.36389867464701336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,7168,0.7305591371324328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,7168,0.6960639953613281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,5120,0.32077333662245006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,6144,0.6039715343051487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,4096,0.25316890080769855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,5120,0.5038133197360569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,6144,0.6128142144944932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,5120,0.5087671279907227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,3584,0.2161475552452935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,4096,0.3985244433085124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,3072,0.19154844019148085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,4096,0.4085635609096951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,3584,0.3452400101555719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,3584,0.3602044317457411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,2560,0.1608311070336236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,3072,0.3029395474327935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,3072,0.31329067548116046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,2048,0.1314257780710856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,2560,0.26627643903096515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,2560,0.2536657651265462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,2048,0.20803822411431205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,1536,0.11343910959031846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,2048,0.22192977534400093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,1024,0.07773866918351915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,1536,0.17786133289337158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,1536,0.1646720038519965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,1024,0.134662217564053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,768,0.05925866630342272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,512,0.044842667049831815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,768,0.11283111572265625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,768,0.09871911340289646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,512,0.08154933320151435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,256,0.033203555477990046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,512,0.09063288900587295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,128,0.029069334268569946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,256,0.0701191094186571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,256,0.07124622000588311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,64,0.024412444896168176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,128,0.06535733408398099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,16384,128,0.05589689148796929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,16384,32,0.024784000383483037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,32,0.06764800018734403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,64,0.06652710835138957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,16384,1024,0.12311732769012451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,65536,3.015999052259657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,51200,2.4552826351589627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,65536,6.3888702392578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,65536,5.334286159939236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,16384,0.7575190862019857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,51200,4.956310272216797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,16384,1.4229422675238714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,12288,0.5943964322408041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,16384,1.3588551415337458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,12288,1.0188933478461373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,10240,0.519055101606581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,12288,1.01328796810574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,8192,0.4233395523495144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,51200,4.173704783121745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,10240,0.8406062126159668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,10240,0.8524586889478895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,7168,0.3543235460917155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,8192,0.6855306625366211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,8192,0.6850070953369141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,6144,0.30129512151082355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,7168,0.580692450205485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,7168,0.6071368853251139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,6144,0.5044124391343858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,5120,0.2559315628475613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,6144,0.525344901614719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,4096,0.20631199412875703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,5120,0.4134533405303955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,5120,0.4425200091467963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,4096,0.3251769012875027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,4096,0.35376977920532227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,3584,0.283644437789917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,3584,0.3111013306511773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,3072,0.15513422754075792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,3072,0.24424621793958876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,3072,0.27101243866814506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,2560,0.1250800026787652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,3584,0.1758648951848348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,2560,0.2079119947221544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,2560,0.2300648954179552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,2048,0.17091022597418892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,2048,0.1914302243126763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,1536,0.07971200015809801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,1024,0.05803910891215006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,1536,0.15320089128282335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,1536,0.1347528828514947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,1024,0.10129333204693264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,768,0.04282666577233208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,1024,0.11683466699388291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,512,0.033032000064849854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,768,0.08104266722997029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,768,0.09810222519768609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,256,0.03157777918709649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,2048,0.11147022247314453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,512,0.0669706662495931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,512,0.0787644452518887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,128,0.02201777696609497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,256,0.05696711275312635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,256,0.06144977940453423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,64,0.019208888212839764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,128,0.05272444420390659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,12288,32,0.019871110717455547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,32,0.05435288945833842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,12288,64,0.05346133311589559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,12288,128,0.04805244339836968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,65536,2.7554577721489797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,51200,2.149956385294596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,65536,5.651932610405816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,65536,4.80265850490994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,16384,0.6711546579996744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,51200,4.276387532552083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,12288,0.5290773179796007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,16384,1.253544913397895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,51200,3.7723901536729603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,16384,1.226618660820855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,10240,0.4424115551842584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,12288,0.9000053405761719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,12288,0.917053328620063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,10240,0.7435422473483615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,8192,0.3396408822801378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,7168,0.3030800024668376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,10240,0.7697857750786675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,8192,0.5879973305596246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,8192,0.6229653358459473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,7168,0.5223866568671333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,6144,0.26593067910936147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,5120,0.20267289214664033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,6144,0.4430524508158366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,6144,0.4742470847235785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,4096,0.16932711336347792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,5120,0.37062666151258683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,5120,0.4012097782558865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,3584,0.14940178394317627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,4096,0.2906488842434353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,4096,0.32065243191189235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,7168,0.5485075314839681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,3584,0.2525751060909695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,3584,0.2811831103430854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,2560,0.10768977801005046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,3072,0.21744710869259307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,3072,0.24518577257792154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,2048,0.08966577715343899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,2560,0.18558400207095674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,2560,0.20795822143554688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,1536,0.07328533464007907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,2048,0.15239200327131483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,2048,0.17338132858276367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,1024,0.04678222205903795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,1536,0.12018399768405491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,1536,0.13950844605763754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,3072,0.12842133310106066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,768,0.04022577736112807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,1024,0.09035466776953803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,512,0.02827289038234287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,1024,0.105877333217197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,768,0.08905333280563354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,256,0.021951110826598272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,512,0.0597662197219001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,512,0.07150844732920329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,768,0.07284710804621379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,128,0.01959911154376136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,256,0.05004177822007073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,256,0.05603377686606514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,64,0.017599110802014668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,128,0.04639644424120585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,10240,128,0.043888890080981784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,10240,32,0.018248000078731112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,64,0.04625599914126926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,10240,32,0.04701155424118042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,65536,2.0838036007351346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,51200,1.6182809405856664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,65536,4.700965457492404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,16384,0.5708755387200249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,65536,4.407460530598958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,16384,1.0838026470608182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,16384,1.123600853814019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,12288,0.4071146647135417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,51200,3.76945071750217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,12288,0.7830702463785807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,10240,0.3534631199306912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,12288,0.8413679864671496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,51200,3.4434774186876087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,8192,0.27214309904310435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,10240,0.6493155691358778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,7168,0.2335448794894748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,10240,0.7354444397820367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,8192,0.5164560212029351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,8192,0.5712871021694607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,6144,0.19707466496361625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,7168,0.48072444068060977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,5120,0.1678711175918579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,7168,0.5034969117906358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,6144,0.39008710119459367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,4096,0.12840178277757433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,5120,0.3246844344668918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,5120,0.36843111779954696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,4096,0.25442666477627224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,4096,0.2936177783542209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,3584,0.22162044048309326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,3072,0.0976746678352356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,3584,0.2583048873477512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,3072,0.1910711129506429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,3072,0.22478932804531523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,2560,0.08389777607387966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,6144,0.43697865804036456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,2048,0.06880533032947116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,3584,0.11909866333007812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,2560,0.16266932752397326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,2560,0.19018222226036918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,1536,0.05304888884226481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,2048,0.15839021735721165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,2048,0.13401688469780818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,1536,0.10557866758770412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,1024,0.03976800044377645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,1536,0.12747377819485134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,768,0.03103999959097968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,1024,0.09730844365225898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,512,0.02374222212367588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,768,0.08163822359508939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,768,0.06382311052746244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,1024,0.07968533039093018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,256,0.01826400061448415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,512,0.06541511085298327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,512,0.05216888917817009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,128,0.017528888252046373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,256,0.050628443559010826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,64,0.018605333235528734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,128,0.03794044587347243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,8192,128,0.04044533438152737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,256,0.042488889561759106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,8192,32,0.01923466722170512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,64,0.0373911095990075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,8192,32,0.03698933455679152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,51200,1.3648026784261067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,65536,1.9907031589084203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,65536,4.325381384955512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,16384,0.4621200031704373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,65536,4.143522474500868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,51200,3.468457751803928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,12288,0.34059998724195695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,51200,3.2392230563693576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,16384,1.0014240476820204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,10240,0.2775813208685981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,12288,0.7286213239034017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,16384,1.057407061258952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,8192,0.24180799060397676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,12288,0.794015990363227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,10240,0.6027039951748318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,10240,0.6656391355726454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,7168,0.19493156009250215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,8192,0.47976091172960067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,6144,0.1743182208802965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,8192,0.5393990940517849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,7168,0.4222782187991672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,7168,0.47452979617648655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,5120,0.14071288373735216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,6144,0.362956444422404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,4096,0.11218755775027806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,6144,0.4122009012434218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,5120,0.30184711350335014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,5120,0.347471104727851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,4096,0.2370391156938341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,3584,0.10473333464728461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,4096,0.2765875657399495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,3072,0.08846133285098606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,3584,0.20555199517144096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,3584,0.2427511215209961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,2560,0.07683200306362577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,3072,0.21109422047932944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,3072,0.17838311195373535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,2048,0.05911733044518364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,2560,0.15161066585116917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,2560,0.17976889345380995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,1536,0.0507919987042745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,2048,0.14907644854651556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,2048,0.12468800279829238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,1024,0.03429155548413595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,1536,0.0986515548494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,1536,0.11988622612423366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,768,0.028584887584050495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,1024,0.09135999944474961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,1024,0.07419377565383911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,768,0.05940000216166178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,512,0.019542222221692402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,768,0.07725689146253797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,256,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,512,0.047826667626698814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,512,0.06208178069856432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,128,0.014121777481502958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,256,0.048057778014077075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,256,0.03770844472779168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,64,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,7168,128,0.03830400109291077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,128,0.03498222099410163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,7168,32,0.012263110942310758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,64,0.032960888412263654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,7168,32,0.0329751107427809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,65536,1.6487520005967882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,51200,1.3441137737698023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,65536,4.069001091851129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,16384,0.4281280040740967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,51200,3.159213383992513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,65536,4.003454420301649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,12288,0.33275821473863393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,16384,0.9253430896335177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,51200,3.1333963606092663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,16384,1.0230951309204102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,10240,0.2784915500217014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,12288,0.7684355311923556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,10240,0.5562675793965658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,8192,0.22179199589623344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,10240,0.6433102289835612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,7168,0.1882693370183309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,8192,0.4428951210445828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,12288,0.6811333232455783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,8192,0.5217884381612142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,6144,0.15254933304256862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,7168,0.39118488629659015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,7168,0.45929691526624894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,5120,0.12183022499084473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,6144,0.336317327287462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,4096,0.0995111134317186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,6144,0.3981911076439752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,5120,0.2795822090572781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,5120,0.3350577884250217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,3584,0.08826044532987808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,4096,0.21891644265916613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,3072,0.07661688990063138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,3584,0.19050399462381998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,4096,0.26738667488098145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,2560,0.06569155719545153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,3072,0.16334756215413412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,3072,0.20281955930921766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,2048,0.05728444125917223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,2560,0.17304089334275988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,2560,0.14010222752888998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,1536,0.042167110575569995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,2048,0.11533866988288032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,2048,0.14407644006941053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,1024,0.03199199835459391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,1536,0.09133333630032009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,1536,0.11564889219072129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,768,0.02331466641690996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,1024,0.06812889046139188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,3584,0.2343671056959364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,1024,0.0883822242418925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,512,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,768,0.05406666795412699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,768,0.07442133294211493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,256,0.01443911095460256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,512,0.059063110086652965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,512,0.04300711221165127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,256,0.046141333050198026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,128,0.013405332962671915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,64,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,6144,128,0.03606222073237101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,128,0.03071199854214986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,6144,32,0.01203555530971951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,64,0.029823111163245306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,32,0.029860443539089624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,6144,256,0.03331733412212796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,65536,1.3889200422498915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,51200,1.048424932691786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,65536,3.6955218844943576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,65536,3.741230010986328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,51200,2.8199030558268228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,16384,0.35965243975321454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,16384,0.8388355573018392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,12288,0.25350843535529244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,16384,0.9554915957980685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,51200,2.931087917751736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,10240,0.22674844000074598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,12288,0.6123421986897787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,8192,0.18702933523390028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,12288,0.7202053070068359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,10240,0.5104826821221246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,7168,0.1698755555682712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,8192,0.40824267599317765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,8192,0.4892391098870172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,7168,0.359844446182251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,6144,0.13636088371276855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,7168,0.43096176783243817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,5120,0.11503733528984918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,6144,0.30930399894714355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,6144,0.3734462261199951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,4096,0.0934400028652615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,5120,0.257604440053304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,10240,0.6040142377217611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,5120,0.31382399135165745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,3584,0.08374844657050239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,4096,0.2013404501809014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,4096,0.2498506704966227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,3072,0.07071289088990954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,3584,0.17490666442447236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,2560,0.05921510855356852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,3584,0.22012000613742402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,3072,0.1494284470876058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,2048,0.047873778475655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,3072,0.19027554988861084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,2560,0.12678666909535727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,1536,0.03879200087653266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,2048,0.10589333375295003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,2048,0.13475555843777126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,1024,0.027052443888452318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,1536,0.08379733562469482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,1536,0.10830755366219415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,768,0.020998222960366145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,1024,0.0617191129260593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,1024,0.08230311340755887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,512,0.0162808895111084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,768,0.048998223410712353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,768,0.0697608921262953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,256,0.012772444221708508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,512,0.03801688883039687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,512,0.055825776524013944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,128,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,256,0.029902223083708022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,256,0.040800889333089195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,64,0.009319999979601966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,128,0.027100443840026855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,128,0.032588445478015475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,5120,2560,0.16191555394066706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,5120,32,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,64,0.026759111218982275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,5120,32,0.026476444469557867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,51200,0.8615937762790256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,65536,1.1788203981187608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,65536,3.228888829549154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,16384,0.29728444417317706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,51200,2.584220462375217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,65536,3.4792018466525607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,51200,2.7206204732259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,12288,0.21931110488043892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,16384,0.7570497724745008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,10240,0.1837306684917874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,16384,0.8891235987345377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,12288,0.5578222274780273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,12288,0.6705244382222494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,8192,0.14771289295620388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,10240,0.4648977915445964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,7168,0.12663643889957005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,10240,0.5628684361775717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,8192,0.3724186685350206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,6144,0.11337333255343968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,8192,0.4561991161770291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,7168,0.3279404375288221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,5120,0.0905146665043301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,7168,0.40260712305704754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,6144,0.2826968828837077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,4096,0.07499288850360446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,6144,0.34893332587348086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,5120,0.23523555861579049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,5120,0.29316888915167916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,3584,0.06246311134762234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,4096,0.18329244189792207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,3072,0.05311288767390781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,4096,0.2325457731882731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,3584,0.15908443927764893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,3584,0.20351021819644502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,2560,0.04943111207750109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,3072,0.13602222336663142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,3072,0.17684622605641684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,2560,0.11516622702280681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,2048,0.038274665673573814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,2560,0.1504017776913113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,1536,0.030935111973020766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,2048,0.09437866343392266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,2048,0.1254097753100925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,1024,0.02172800070709652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,1536,0.07593955596288045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,1024,0.07649688588248359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,1024,0.055200000603993736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,768,0.017444445027245414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,768,0.06469866964552137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,768,0.04392888810899523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,512,0.013044444223244986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,512,0.051274667183558144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,256,0.010442666709423065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,512,0.03472977876663208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,256,0.026867555247412786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,1536,0.10104711188210382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,256,0.04020622372627258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,128,0.009328000247478485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,64,0.009350222018029954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,128,0.02407022151682112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,4096,32,0.009663110805882348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,64,0.023690667417314317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,4096,128,0.0317466656366984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,4096,32,0.023697778582572937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,65536,1.0304062101576064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,51200,0.8032817840576172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,65536,2.9673343234592013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,16384,0.2770177788204617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,65536,3.342950185139974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,51200,2.380185869004991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,16384,0.7050791316562228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,12288,0.21065778202480742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,51200,2.621241675482856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,12288,0.5260986751980251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,16384,0.8554097811381022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,10240,0.14848355452219644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,12288,0.6469066407945421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,8192,0.12612356079949275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,10240,0.44109511375427246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,10240,0.5421057807074653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,8192,0.3543813228607178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,7168,0.10504800081253052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,8192,0.43908177481757266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,6144,0.08819466829299927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,7168,0.3124177720811632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,5120,0.07567022244135539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,6144,0.2688213454352485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,7168,0.3879351086086697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,6144,0.3366266621483697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,4096,0.06102310948901706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,5120,0.22291554345024955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,3584,0.05484444565243191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,5120,0.2822391192118327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,4096,0.1758657826317681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,4096,0.22429688771565756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,3072,0.04519288738568624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,3584,0.1518097718556722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,3584,0.19666488965352377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,2560,0.038626667526033186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,3072,0.1296506722768148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,2048,0.03398133317629496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,2560,0.10931555430094402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,2560,0.14460355705685085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,1536,0.025719111164410908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,2048,0.09167822202046712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,2048,0.12206578254699707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,1024,0.018831110662884183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,1536,0.07264000177383423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,1536,0.09933955801857842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,1024,0.053455998500188194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,768,0.015663999650213454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,1024,0.07414399915271334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,512,0.011983999775515662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,768,0.041780445310804576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,768,0.0637048880259196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,256,0.00943199959066179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,512,0.032673776149749756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,512,0.052523556682798594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,128,0.00795911086930169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,3072,0.17045866118537056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,256,0.025283555189768474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,256,0.039022223816977605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,64,0.007991111112965478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3584,32,0.008345777789751688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,128,0.022308443983395893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3584,128,0.03095199995570713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,64,0.022287999590237934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3584,32,0.022042666872342426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,65536,0.8354489008585612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,51200,0.669058640797933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,65536,2.7656898498535156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,16384,0.215759105152554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,65536,3.3384674919976125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,51200,2.616355472140842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,16384,0.663256910112169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,12288,0.17463110552893746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,51200,2.209066602918837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,12288,0.4984746509128147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,16384,0.8537679778205024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,10240,0.14184178246392146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,12288,0.6447368727789985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,10240,0.4169635507795546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,8192,0.12415733602311875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,10240,0.5416800181070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,7168,0.1023706661330329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,8192,0.3362915515899658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,8192,0.4378053347269694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,7168,0.2964506679111057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,6144,0.08798133002387153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,7168,0.3862888813018799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,5120,0.06472266382641263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,6144,0.2549582322438558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,6144,0.33417243427700466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,4096,0.05455911159515381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,5120,0.21203555001152885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,5120,0.2810640070173475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,4096,0.16651554902394614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,3584,0.04653066727850172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,4096,0.2230773237016466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,3072,0.04007555709944831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,3584,0.14393777317470974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,3584,0.19569599628448486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,2560,0.036892443895339966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,3072,0.12378488646613227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,3072,0.1689128875732422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,2048,0.03011200163099501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,2560,0.10444711314307319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,2560,0.14389333460066053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,1536,0.02281066609753503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,2048,0.0851057767868042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,2048,0.12014844682481553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,1024,0.017595556047227647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,1536,0.06501955456203885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,1536,0.0961315565639072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,768,0.014179555906189812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,1024,0.04829155405362447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,1024,0.07228977812661065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,512,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,768,0.03914577762285868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,768,0.06160889069239298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,256,0.010130667024188572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,512,0.04714311162630717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,256,0.023027555810080633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,256,0.036285334163241915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,128,0.00998755544424057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,128,0.020959110723601446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,3072,128,0.029888000753190782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,64,0.020568889048364427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,32,0.009700444837411245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,32,0.020406223005718656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,3072,512,0.030424889591005113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,3072,64,0.009330666727489894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,65536,0.7587111261155871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,51200,0.5961253378126357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,65536,2.541614320543077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,16384,0.19334399700164795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,51200,2.043440077039931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,16384,0.6267164548238119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,12288,0.1461662186516656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,65536,3.21014404296875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,12288,0.47256888283623594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,16384,0.821706665886773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,10240,0.10857688718371922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,51200,2.512019475301107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,8192,0.10054400232103135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,12288,0.620664914449056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,10240,0.39644355244106716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,7168,0.08156622118420072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,10240,0.5208977593315972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,8192,0.31946754455566406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,6144,0.0690444442960951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,8192,0.42170045110914445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,7168,0.2812311119503445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,5120,0.05722666449016995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,7168,0.3722444375356038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,6144,0.2429137759738498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,6144,0.3220213254292806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,5120,0.2008533345328437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,5120,0.27099643813239205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,3584,0.03917422228389316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,4096,0.15753866566552058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,4096,0.21495910485585532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,3072,0.03438222077157762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,3584,0.136409772766961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,3584,0.18822133541107178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,2560,0.029538667864269678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,3072,0.1163982285393609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,3072,0.16217688719431558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,2048,0.02532711128393809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,2560,0.09815644555621678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,2560,0.13915022214253744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,1536,0.02036888897418976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,2048,0.08000000317891438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,4096,0.04457066787613762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,2048,0.11540444691975911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,1024,0.014515555567211576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,1536,0.0610426664352417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,1536,0.09295466873380874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,768,0.011494222614500256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,1024,0.0457911127143436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,1024,0.07019733058081733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,512,0.009352000223265754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,768,0.036908444431093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,768,0.059761775864495166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,256,0.007745777567227681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,512,0.027963555521435205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,256,0.03556888964441087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,256,0.022084444761276245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,128,0.006736889067623351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,128,0.02886311213175456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,64,0.00629333323902554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,128,0.019952888290087383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2560,32,0.006682666639486949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,64,0.019036443697081674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2560,32,0.019316444794336956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2560,512,0.04561866654290093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,65536,0.5582568910386827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,51200,0.447737799750434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,65536,2.381327099270291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,16384,0.15357422828674316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,51200,1.8695013258192275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,65536,3.066136042277018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,16384,0.5930630895826552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,12288,0.10847644673453437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,16384,0.7865066528320312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,10240,0.10434489117728339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,12288,0.44710667928059894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,12288,0.5961990886264378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,8192,0.08799555566575791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,10240,0.3746364381578233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,51200,2.4025724199083114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,10240,0.5018355581495497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,7168,0.07034489181306627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,8192,0.3022017743852404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,6144,0.06383733616934882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,8192,0.40632088979085285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,7168,0.26671910285949707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,7168,0.35779645707872176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,6144,0.22861154874165854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,5120,0.05183733171886868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,6144,0.3100275463528103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,4096,0.04071822100215488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,5120,0.19061156113942465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,5120,0.2605982091691759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,3584,0.033511112133661904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,4096,0.1489137808481852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,3072,0.029363556040657893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,4096,0.20539644029405382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,3584,0.12912266784244114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,3584,0.1805244419309828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,2560,0.02646044393380483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,3072,0.11024177736706203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,3072,0.15585245026482478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,2048,0.02307644486427307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,2560,0.09232799874411689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,2560,0.13274666998121473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,1536,0.017821333474583097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,2048,0.11093778080410427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,1024,0.012792000340090858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,1536,0.05746044715245565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,1536,0.08891377846399943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,768,0.011202666494581433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,1024,0.04049688908788893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,1024,0.06613955232832167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,512,0.009043555292818282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,768,0.03386844529045953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,768,0.05550666650136312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,256,0.007664889097213745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,512,0.02650044361750285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,512,0.043881777260038585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,128,0.007280888656775157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,256,0.02030666669209798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,128,0.017924444542990792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,64,0.006973333656787872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,128,0.027466666367318895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,2048,32,0.007172444628344641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,2048,0.07633333073721991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,64,0.017522666189405654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,2048,32,0.017270222306251526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,65536,0.41620800230238175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,51200,0.3303448888990614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,2048,256,0.033655110332700945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,65536,2.2164427439371743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,16384,0.11479200257195367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,51200,1.7408826616075304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,16384,0.5563431315951878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,12288,0.08401688602235581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,65536,2.9390640258789062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,51200,2.3002808888753257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,10240,0.08213777674569024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,16384,0.7543795373704699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,12288,0.420634667078654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,8192,0.06834666596518622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,10240,0.3533528910742865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,12288,0.5721208784315321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,10240,0.48151111602783203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,8192,0.2849404546949598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,7168,0.057032002343071826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,8192,0.389158222410414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,6144,0.04766755633884006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,7168,0.2508017751905653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,6144,0.21595111158159044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,7168,0.3437919881608751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,5120,0.03974044322967529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,6144,0.29666932423909503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,4096,0.032469332218170166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,5120,0.249400880601671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,5120,0.18021510707007515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,4096,0.140864888827006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,3584,0.02844711144765218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,4096,0.19668000274234346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,3072,0.024433778391944036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,3584,0.12159910466935898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,3584,0.1729280021455553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,3072,0.10404266913731892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,2560,0.021320889393488567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,3072,0.14848000473446316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,2048,0.017981333865059745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,2560,0.08725066979726155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,2560,0.12756533092922634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,2048,0.07152533531188965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,1536,0.014289778139856128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,2048,0.10649689038594563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,1024,0.010679999987284342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,1536,0.05516889029079013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,1536,0.08569422033098008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,768,0.009161777794361115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,1024,0.038422223594453596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,1024,0.06240799691942003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,512,0.007975111405054728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,768,0.05311288767390781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,768,0.03155644403563605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,256,0.0069066666894488865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,512,0.025016888976097107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,512,0.0418124430709415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,128,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,256,0.03309866786003113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,256,0.0192577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,64,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,128,0.016917333006858826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1536,128,0.026518222358491685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1536,32,0.005944889038801193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,64,0.01650311052799225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1536,32,0.01583377851380242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,65536,0.29413776927524143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,51200,0.2253662215338813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,51200,1.606536865234375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,16384,0.07879555225372314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,65536,2.0488444434271917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,16384,0.5216524336073134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,51200,2.2001164754231772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,12288,0.05893511242336697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,16384,0.7222053209940592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,65536,2.81330935160319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,10240,0.050981332858403526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,12288,0.3939715491400824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,8192,0.04039288891686334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,12288,0.5480408668518066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,10240,0.3309413327111138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,10240,0.4607226583692763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,8192,0.26708798938327366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,7168,0.03578844335344102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,6144,0.03125422199567159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,8192,0.3725751241048177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,7168,0.2349697748819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,5120,0.026434666580624048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,6144,0.20308355490366617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,6144,0.2840346760219998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,4096,0.02218755582968394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,5120,0.1685911152097914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,5120,0.2397786776224772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,4096,0.1318568918440077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,3584,0.020364445116784837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,4096,0.1888951063156128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,3072,0.017940445078743827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,3584,0.11322132746378581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,3584,0.16527556048499212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,2560,0.01629688839117686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,3072,0.09619555870691936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,3072,0.1426408953136868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,2048,0.013070222404268054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,7168,0.32951556311713326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,2560,0.08156177732679579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,2560,0.12184622552659775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,1536,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,2048,0.06665866904788546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,2048,0.10103288624021743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,1024,0.00831911133395301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,1536,0.05040711164474487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,1536,0.08164444234636095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,768,0.007342221836249034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,1024,0.03546311126814948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,1024,0.05955644448598226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,512,0.006261333409282897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,768,0.05033955640263028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,768,0.029453333881166246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,256,0.0052791109515561005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,512,0.040355556541019015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,256,0.03118844495879279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,256,0.017528888252046373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,128,0.004624888714816836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,1024,128,0.025038222471872967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,128,0.01555999947918786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,64,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,1024,32,0.004949333353175057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,32,0.014549333188268872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,64,0.015091554986106025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,65536,0.23608978589375815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,1024,512,0.023031110564867657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,51200,0.18070667319827607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,65536,2.0018853081597223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,51200,1.5643048816257055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,16384,0.06101510922114054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,16384,0.5088693300882975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,12288,0.04607466525501675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,65536,2.8102364010281033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,12288,0.38498756620619035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,16384,0.7218595610724555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,10240,0.03962755534383986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,12288,0.5475040011935765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,8192,0.03167111012670729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,10240,0.32343554496765137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,10240,0.45960532294379336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,7168,0.028243554963005915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,8192,0.26056710879007977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,51200,2.2036914825439453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,7168,0.2292391194237603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,8192,0.3733564482794867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,6144,0.024241778585645888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,7168,0.3283911016252306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,5120,0.020769778225156996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,6144,0.19733066029018828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,4096,0.01754044493039449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,6144,0.2849840058220757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,5120,0.2401804394192166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,4096,0.12895554966396755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,3584,0.01588622232278188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,4096,0.18878844049241808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,3072,0.013807111316257052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,3584,0.11098221937815349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,3072,0.09389955467647976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,3584,0.1652862230936686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,2560,0.01220266686545478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,3072,0.14220710595448813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,2048,0.01033155537313885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,2560,0.1218222247229682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,2048,0.06480711036258273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,5120,0.16353866789076063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,1536,0.008669333325492011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,2560,0.07960444688796997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,1024,0.007367110914654202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,2048,0.10092888938056098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,1536,0.04876088764932421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,1536,0.08089866903093126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,768,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,1024,0.05900355842378405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,768,0.028147555059856836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,512,0.005640000104904175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,768,0.0499946673711141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,1024,0.03411022159788344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,256,0.004881777697139316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,512,0.022554667459593877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,128,0.0041786668201287585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,256,0.03091999888420105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,256,0.016921778519948322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,128,0.01478844384352366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,128,0.025045333637131587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,64,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,768,32,0.005605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,64,0.014114666316244336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,65536,0.16965777344173857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,768,32,0.013861333330472311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,51200,0.13535822762383357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,768,512,0.039748443497551814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,65536,1.9187306298149958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,51200,1.5036747190687392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,16384,0.04642755455440945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,16384,0.4903804461161296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,65536,2.6787821451822915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,12288,0.035880890157487653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,16384,0.6885289086235894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,51200,2.1001048617892795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,10240,0.02985422147644891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,12288,0.3711644543541803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,10240,0.31158222092522514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,8192,0.022282666630215112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,10240,0.43958399030897355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,12288,0.523304886288113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,7168,0.019673777951134574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,8192,0.25179023212856716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,6144,0.01737600068251292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,8192,0.35523732503255206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,7168,0.31334045198228627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,5120,0.015116444892353482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,7168,0.22227377361721465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,6144,0.19064088662465414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,6144,0.27116089397006565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,4096,0.012608889076444836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,5120,0.1581831110848321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,5120,0.2281013329823812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,4096,0.12346577644348145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,3584,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,4096,0.17896444267696807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,3072,0.01035377797153261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,3584,0.10683111349741618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,3584,0.15746400091383192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,2560,0.009297777381208207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,3072,0.0908435583114624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,3072,0.13513422012329102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,2048,0.008005333443482717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,2560,0.07543822129567464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,2560,0.11536533302730984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,1536,0.007267555428875818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,2048,0.06171822547912598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,2048,0.09595644474029541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,1536,0.04631111025810242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,1024,0.005921777751710679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,1536,0.07682577768961589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,768,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,1024,0.03259999884499444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,1024,0.05615288681454129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,512,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,768,0.026791999737421673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,768,0.04792177677154541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,256,0.0038613333470291565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,512,0.02145333256986406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,512,0.0380364457766215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,128,0.0038533334930737815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,256,0.015862221519152325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,256,0.02952444553375244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,64,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,128,0.013435555828942193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,512,128,0.023765333824687537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,512,32,0.00350222239891688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,64,0.01313511116637124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,65536,0.11254844400617812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,512,32,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,51200,0.08638133605321248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,51200,1.4451209174262152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,16384,0.03737599982155694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,65536,1.8458177778455946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,16384,0.47264353434244794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,12288,0.027292443646325007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,16384,0.6852693557739258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,65536,2.666471905178494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,10240,0.020054222808943856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,12288,0.3575884501139323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,12288,0.5204159948560927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,10240,0.29966311984592015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,8192,0.016563556260532804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,10240,0.4367057747311062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,7168,0.014477334088749357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,8192,0.24200177192687988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,51200,2.091607199774848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,8192,0.35352712207370335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,7168,0.21313956048753527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,6144,0.013361777696344586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,5120,0.011588444312413534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,7168,0.31159999635484487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,6144,0.18335555659400093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,4096,0.009675555759006077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,6144,0.26939911312527126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,5120,0.15321511692470974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,5120,0.22608889473809135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,4096,0.11863999896579319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,3584,0.008979555633332994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,4096,0.17808977762858072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,3072,0.008332444561852349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,3584,0.10252266459994847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,2560,0.007324444750944774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,3584,0.1564942200978597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,3072,0.08656799793243408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,2048,0.0063573333124319715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,3072,0.1351591083738539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,2560,0.07249955336252849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,2560,0.11419910854763454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,1536,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,2048,0.05926666657129923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,1024,0.00489066665371259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,2048,0.09499733315573798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,1536,0.04370222157902188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,1536,0.07612178060743544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,768,0.004489777816666497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,1024,0.03091644578509861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,512,0.003941333128346337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,1024,0.05600621965196398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,768,0.02568977740075853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,768,0.047284444173177086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,256,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,512,0.020634666085243225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,128,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,512,0.037821332613627114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,256,0.015072888798183866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,64,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,128,0.013086222112178802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,128,0.02340000040001339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,64,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,256,32,0.0033733333564466904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,256,32,0.01237066669596566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,256,256,0.02886311213175456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,65536,0.10052088896433513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,51200,0.07395999961429171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,65536,1.8089165157741969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,51200,1.4164639578925238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,16384,0.029527111185921565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,65536,2.6657305823432074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,16384,0.4629813300238715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,12288,0.020557334025700886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,51200,2.089946746826172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,10240,0.015068444940778943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,16384,0.6844071282280816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,12288,0.35042312410142684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,8192,0.012503999802801343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,12288,0.5195715692308214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,10240,0.2937555578019884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,7168,0.011189333266682096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,10240,0.43645866711934406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,8192,0.2376533349355062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,8192,0.3530471060011122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,6144,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,7168,0.2090382178624471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,5120,0.008996444443861643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,7168,0.31123556031121147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,6144,0.17988533443874785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,4096,0.007732444339328342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,5120,0.1489351060655382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,6144,0.26954221725463867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,5120,0.22627109951443145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,3584,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,4096,0.11618755923377143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,3072,0.006626666833957036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,4096,0.17762577533721924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,3584,0.10007733106613159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,2560,0.006207110981146495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,3584,0.15608978271484375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,3072,0.0848640004793803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,3072,0.13423111703660753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,2048,0.005551110953092575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,2560,0.07131733496983846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,1536,0.004905777672926585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,2048,0.057167112827301025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,2048,0.09482844670613606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,2560,0.11446489228142633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,1024,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,1536,0.04246133234765795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,1536,0.07591111130184598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,768,0.0038693332009845306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,1024,0.0551546679602729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,1024,0.03055644366476271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,512,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,768,0.046560890144772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,768,0.025207999679777358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,256,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,512,0.03737599982155694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,512,0.020440000626775954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,128,0.0031475555151700974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,256,0.014509333504570855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,256,0.028789334826999243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,64,0.002973333415057924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4096,128,128,0.022699554761250813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,128,0.012591111163298288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,64,0.012058667010731168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,128,32,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,51200,0.07059022453096178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,128,32,0.0031439999325407874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,16384,0.028078221612506445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,65536,0.09586933586332534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,12288,0.016734222571055096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,51200,1.4143235948350694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,10240,0.014152889450391134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,16384,0.4626604186164008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,8192,0.012077333198653327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,65536,1.8070231543646917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,7168,0.011035555766688453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,12288,0.3503359953562419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,10240,0.29362932840983075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,8192,0.2369635634952121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,5120,0.008655110994974772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,7168,0.20840356085035536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,4096,0.007612444460391998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,6144,0.17988622188568115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,3584,0.006909333169460297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,5120,0.1485662195417616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,4096,0.11602932877010769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,3072,0.006656000183688269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,2560,0.0058986664646201665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,3584,0.10000355376137628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,2048,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,2560,0.07046933306588067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,1536,0.004603555632962121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,2048,0.0573635564910041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,1024,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,1536,0.04173955652448866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,768,0.0038622220357259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,6144,0.010009778042634329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,1024,0.03054222133424547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,512,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,768,0.024920889072948035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,256,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,512,0.019928889142142404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,128,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,256,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,64,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,128,0.012375999655988483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,64,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,64,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,65536,0.09024533298280503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,32,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,51200,0.07117778062820435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,64,3072,0.08448533217112224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,16384,0.02552177839808994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,51200,1.4145636028713648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,12288,0.013400889105266996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,16384,0.4623564614189996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,65536,1.8081732855902777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,10240,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,8192,0.009778666827413771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,12288,0.349218659930759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,7168,0.009164444274372524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,10240,0.29327111774020725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,8192,0.2368248833550347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,6144,0.00869066682126787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,5120,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,7168,0.2084897756576538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,6144,0.1789564424090915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,4096,0.00963466697269016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,3584,0.008640888664457533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,5120,0.14835555023617214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,3072,0.00795555528667238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,2560,0.007256000406212277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,4096,0.11618133385976155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,2048,0.005233777893914117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,3584,0.10025688674714829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,1536,0.004581333448489507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,2560,0.07027466429604425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,1024,0.003956444561481476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,2048,0.057667553424835205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,768,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,1536,0.041730665498309664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,512,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,3072,0.08519555462731256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,1024,0.030208001534144085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,256,0.003306666596068276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,768,0.0247697780529658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,128,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,512,0.01990133358372582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,256,0.014441778262456259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4096,32,32,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,128,0.012040889097584618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,64,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4096,32,32,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,16384,4.07621341281467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,16384,5.599826812744141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,12288,3.0643094380696616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,12288,3.8660702175564237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,10240,2.6794613732231984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,8192,2.2663538191053605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,51200,12.894558376736112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,10240,3.1501280466715493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,7168,2.115797254774306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,8192,2.417706595526801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,7168,2.1195670233832464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,6144,1.586314625210232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,5120,1.279256926642524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,6144,1.795946756998698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,4096,1.0119270748562283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,4096,1.1790986590915256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,5120,1.4997200436062281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,3584,0.8898631201850044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,3584,1.0334275563557942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,3072,0.854282644059923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,3072,0.8859493467542859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,51200,19.54498969184028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,2560,0.5987520217895508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,2048,0.48227998945448136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,2560,0.7565600077311198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,1536,0.38875201013353133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,2048,0.6198284361097548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,1024,0.28135021527608234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,768,0.23445333374871147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,1536,0.48787556754218203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,512,0.15529333220587835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,1024,0.3668951193491618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,256,0.12736445003085667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,768,0.3027733431922065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,128,0.10466488864686753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,512,0.24754577212863496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,256,0.22117422686682808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,64,0.08655999766455756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,65536,32,0.09233866797553168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,128,0.20900621679094103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,64,0.21198577351040312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,65536,32,0.21620355712042916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,51200,10.309522840711805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,65536,13.259139166937935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,51200,15.566757202148438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,16384,3.3634630839029946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,12288,2.492980533175998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,16384,4.435829162597656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,65536,19.31656222873264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,12288,3.102559195624458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,10240,2.0899821387396917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,8192,1.6712835099962022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,10240,2.5168213314480252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,7168,1.5240977605183919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,7168,1.6956986321343317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,8192,1.9340623219807942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,6144,1.432992935180664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,5120,1.0228062735663521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,6144,1.2190870708889432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,4096,0.8351528909471301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,5120,1.199853367275662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,3584,0.7011635568406848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,3072,0.6228488816155328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,4096,0.9436213175455729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,3584,0.8394159740871854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,2048,0.40894222259521484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,2560,0.5291155709160699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,3072,0.7098239792717828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,1536,0.3290986749860975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,2560,0.6074675454033746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,2048,0.4964195357428656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,1536,0.3906213442484538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,1024,0.22799645529852974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,512,0.14407644006941053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,768,0.18810578187306723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,1024,0.29534133275349933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,256,0.0987244447072347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,768,0.23350845442877877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,128,0.08220177888870239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,512,0.197723560863071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,256,0.17678133646647134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,64,0.07175911135143705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,128,0.1660835610495673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,51200,32,0.07567555374569364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,64,0.1687706708908081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,51200,32,0.1716808875401815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,51200,3.0520284440782337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,65536,3.9332487318250866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,16384,1.0220035976833768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,12288,0.7802577548556857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,16384,1.6624622344970703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,51200,5.851367950439453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,12288,1.188725365532769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,10240,0.6126426590813531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,8192,0.5236079957750108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,8192,0.7648320198059082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,10240,0.9745457967122396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,7168,0.47858312394883895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,65536,7.288068135579427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,6144,0.37651822302076554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,5120,0.3461742136213515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,7168,0.671928882598877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,6144,0.5741662449306911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,5120,0.4750257598029242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,3584,0.22034755018022326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,4096,0.37340444988674587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,3072,0.21126222610473633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,3584,0.3264168898264567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,2560,0.16190932856665716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,3072,0.2817715538872613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,2048,0.13061511516571045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,1536,0.10026755597856309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,2560,0.24063555399576822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,1024,0.07570755481719971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,2048,0.19822843869527182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,768,0.055868446826934814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,1536,0.15577156013912624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,512,0.043338666359583534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,1024,0.11735999584197998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,256,0.032807999187045626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,768,0.09479733308156331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,128,0.028057777219348486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,512,0.07901155286365084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,64,0.0240328891409768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,256,0.06758577956093682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,32,0.024887111451890733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,128,0.06285333633422852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,64,0.06373688909742567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,16384,32,0.06424178017510308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,16384,4096,0.27260979016621906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,51200,2.399115456475152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,65536,2.9167217678493924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,16384,0.7773502137925891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,12288,0.573222213321262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,16384,1.3409724765353734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,10240,0.5056959788004557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,12288,0.9609751171535916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,51200,4.641790178087023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,8192,0.4447333547804091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,10240,0.806146674686008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,7168,0.37235556708441836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,65536,5.870628356933594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,8192,0.6271733178032769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,6144,0.3268248769972059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,7168,0.5506329006618923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,6144,0.4713333447774251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,5120,0.27059022585550946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,4096,0.20075289408365884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,5120,0.3904888894822862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,3584,0.17823289500342476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,4096,0.3297022183736165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,3072,0.15374667114681667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,2560,0.12909332911173502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,3584,0.26818399959140354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,3072,0.23081599341498482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,2048,0.10587200191285874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,1536,0.08163111077414618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,2560,0.19704622692532012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,2048,0.16250488493177626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,1024,0.056905779573652476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,768,0.045075555642445884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,1536,0.12831644217173258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,512,0.03565244542227851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,1024,0.0964026649792989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,256,0.027091556125217017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,768,0.07840622133678861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,128,0.021872889664438035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,512,0.06435022089216444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,64,0.018909333480728995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,128,0.05128711130883959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,12288,32,0.019254222512245178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,64,0.050959110260009766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,32,0.05187822381655375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,12288,256,0.05484177668889364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,65536,2.4365946451822915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,51200,2.0317245059543185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,16384,0.6455021964179145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,16384,1.180339601304796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,65536,5.171019660101996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,12288,0.47278489006890195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,10240,0.4145377741919623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,12288,0.8531555599636502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,8192,0.34083999527825254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,10240,0.7390782038370768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,7168,0.3081093364291721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,51200,4.1509785122341585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,8192,0.5577928755018446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,7168,0.48921065860324436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,5120,0.20601244767506918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,6144,0.42005154821607804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,4096,0.16571466128031412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,5120,0.3475324312845866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,3584,0.1458186705907186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,4096,0.27332888709174263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,3072,0.13235467010074192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,3584,0.23826755417717826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,2560,0.10969333516226874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,2048,0.0936373339758979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,3072,0.20639467239379883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,2560,0.17578311761220297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,1536,0.06813866562313504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,1024,0.049806220663918384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,2048,0.14510844813452825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,1536,0.11433777544233535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,768,0.04112977782885233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,1024,0.08636177910698785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,512,0.03045066528850132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,256,0.022392888863881428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,768,0.06951466533872816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,128,0.018547554810841877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,512,0.05739821990331014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,64,0.01577422188387977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,256,0.048624889718161694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,128,0.04445866743723551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,32,0.01609777741962009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,64,0.043840001026789345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,10240,32,0.043947554296917386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,10240,6144,0.2633048958248562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,51200,1.6293022367689345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,65536,2.0719004737006292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,16384,0.5340879758199056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,12288,0.3880026605394151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,16384,1.0266479916042752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,10240,0.3144293361239963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,12288,0.7424257596333822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,51200,3.512647204928928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,8192,0.2716968854268392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,7168,0.22715910275777182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,10240,0.6152248912387425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,8192,0.48869509167141384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,6144,0.19627200232611763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,7168,0.4421066708034939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,5120,0.16231021616193983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,4096,0.1280719969007704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,6144,0.3694133228725857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,5120,0.305515554216173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,3584,0.11983644962310791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,3072,0.09674044450124104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,4096,0.24020176463656953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,3584,0.2093182272381253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,2560,0.09179733196894328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,2048,0.06921422481536865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,3072,0.18118755022684732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,2560,0.15433422724405924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,1536,0.053060445520612926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,65536,4.615225897894965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,1024,0.03912533322970072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,2048,0.12806667221917048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,768,0.02942311101489597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,1536,0.10037244690789117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,512,0.022823999325434368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,1024,0.07571378019120958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,256,0.01815111107296414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,512,0.049903111325369946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,768,0.06180000305175781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,128,0.015497777197096082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,256,0.04022044605678982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,64,0.013426666458447775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,8192,32,0.013165333204799227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,128,0.03681244452794393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,64,0.03566933340496487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,8192,32,0.03514133228196038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,51200,1.395901362101237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,65536,1.7842879825168187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,16384,0.4495866563585069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,12288,0.3478417661454942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,16384,0.9509653515285916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,10240,0.3115217685699463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,12288,0.6853822072347006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,65536,4.234866672092013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,8192,0.23603200912475586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,7168,0.2041226625442505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,10240,0.5769946840074327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,51200,3.2824605305989585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,8192,0.4539404445224338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,6144,0.17924000157250297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,5120,0.13959377341800266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,7168,0.40369865629408097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,4096,0.11158043808407253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,6144,0.3422604401906331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,3584,0.1000524428155687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,3072,0.08496266603469849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,4096,0.22277156511942545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,5120,0.28547289636400014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,3584,0.1940631071726481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,2560,0.07141155666775174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,2048,0.06126044193903605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,3072,0.1687706708908081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,1536,0.04804799954096476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,1024,0.03407199846373664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,2048,0.11883821752336289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,768,0.0262106673585044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,2560,0.1440346638361613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,1536,0.09381866455078125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,512,0.02033155494266086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,1024,0.0702951086892022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,256,0.015996444556448195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,768,0.056919111145867236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,512,0.04475288920932346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,128,0.014112000664075216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,256,0.03557422094874912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,64,0.01237777786122428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,7168,32,0.012452444268597497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,128,0.03257511059443156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,32,0.03156088789304098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,7168,64,0.03196533189879523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,51200,1.1809679667154949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,65536,1.579182200961643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,16384,0.39620444509718156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,12288,0.29944443702697754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,16384,0.8653804461161295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,10240,0.2432462109459771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,12288,0.6330986552768284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,8192,0.19154755274454752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,51200,2.9528861575656467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,10240,0.5255475574069554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,7168,0.17514399687449136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,8192,0.4200026724073622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,65536,3.7517556084526906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,6144,0.14825956026713052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,7168,0.3699057896931966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,5120,0.12387466430664062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,4096,0.10113600227567886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,6144,0.3179573218027751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,3584,0.08992533551322089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,3072,0.08254844612545438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,4096,0.20616532696617973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,2560,0.06567200024922688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,3584,0.17928622828589547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,5120,0.26433155271742076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,2048,0.05423466695679558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,3072,0.1540622181362576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,1536,0.044750223557154335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,1024,0.030115554730097454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,2560,0.13283467292785645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,768,0.023258666197458904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,1536,0.08623288737403022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,512,0.019303111566437613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,2048,0.11010222302542792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,1024,0.06456622150209215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,256,0.014369777507252164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,768,0.052095111873414784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,512,0.040469331873787776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,256,0.03198488884501987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,64,0.012730666332774691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,128,0.029236445824305218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,32,0.012808889150619507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,64,0.028816001282797918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,6144,32,0.02878311276435852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,6144,128,0.014628445108731588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,65536,1.4321840074327257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,51200,1.1389510896470811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,16384,0.38542845514085555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,16384,0.7931235631306967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,65536,3.4181991153293185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,12288,0.2801484531826443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,10240,0.22700977325439453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,8192,0.1803662247127957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,12288,0.5790293481614854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,51200,2.7111422220865884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,7168,0.15800266795688206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,10240,0.48177157508002383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,6144,0.1363662216398451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,7168,0.33989689085218644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,5120,0.11319466431935628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,8192,0.38585156864590114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,4096,0.09366044733259414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,3584,0.08274577723609076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,6144,0.2933279938167996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,5120,0.2416151099734836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,4096,0.18941067324744332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,3072,0.0747128857506646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,2560,0.06177600224812826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,3584,0.16467555363972983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,2048,0.05187288920084635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,1536,0.04050666756100125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,3072,0.14152088430192736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,1024,0.028551111618677776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,2048,0.10112888945473565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,2560,0.1209448840883043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,1536,0.0794737802611457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,512,0.01721333298418257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,1024,0.058061334821912974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,256,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,512,0.036320888333850436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,768,0.04709688822428385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,128,0.010355555348926121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,256,0.02869510981771681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,64,0.009369778136412302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,32,0.009337777892748514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,64,0.026103999879625108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,128,0.026671111583709717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,5120,768,0.02294755478700002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,65536,0.9568470848931206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,51200,0.7783715460035535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,5120,32,0.02569688856601715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,16384,0.2731520069970025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,12288,0.20054933759901258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,16384,0.7141200171576606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,10240,0.16001510620117188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,12288,0.5575511190626357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,8192,0.12820088863372803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,10240,0.4397848976982965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,7168,0.11443466610378689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,8192,0.35228800773620605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,65536,3.03767458597819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,6144,0.09989777538511489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,5120,0.0830355551507738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,51200,2.427716361151801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,7168,0.31051821178860134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,4096,0.06910666492250231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,3584,0.06343466705746122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,6144,0.2676160070631239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,5120,0.2206195592880249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,4096,0.1724728875690036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,2560,0.045922666788101196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,3072,0.05296622382269966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,3584,0.14989954895443386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,2048,0.038626667526033186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,3072,0.1291653315226237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,1536,0.029954665237002905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,2560,0.10891911056306626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,1024,0.02223199937078688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,2048,0.08934933609432644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,768,0.018647111124462552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,1536,0.07174399826261732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,512,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,1024,0.052184889713923134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,256,0.012425777812798819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,768,0.04287911123699612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,128,0.00866844422287411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,512,0.032953778902689614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,64,0.007706666986147563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,256,0.02541333271397485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,4096,32,0.008009778128729926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,128,0.023405333360036213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,64,0.02335111134582096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,4096,32,0.022670222653283015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,51200,0.708587540520562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,65536,0.9088933732774523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,16384,0.25199911329481334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,12288,0.18641066551208496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,16384,0.674906677669949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,12288,0.49625423219468856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,10240,0.14759555127885607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,51200,2.2591484917534723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,10240,0.4163493315378825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,8192,0.1311217811372545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,7168,0.10447911421457927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,65536,2.8990550571017795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,8192,0.33503021134270566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,6144,0.08978844351238674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,5120,0.07213511069615682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,6144,0.25441598892211914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,4096,0.05729777945412529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,5120,0.20972355206807455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,3584,0.050160000721613564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,4096,0.16428266631232366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,3072,0.046021332343419395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,3584,0.14260178142123753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,2560,0.037625776396857366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,3072,0.12359021769629584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,2048,0.030581331915325586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,2560,0.10395999749501546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,1536,0.025361778007613286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,7168,0.29960089259677464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,1024,0.018449778358141582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,2048,0.08528000116348267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,768,0.015117333994971381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,1536,0.06811466481950548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,512,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,1024,0.04876266585456001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,256,0.009003555609120263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,768,0.040143112341562905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,512,0.03017066584693061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,128,0.008644444247086843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,256,0.023718222975730896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,64,0.007612444460391998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3584,32,0.008008889026112026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,128,0.021943999661339655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,64,0.021287999219364587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3584,32,0.022037333912319605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,51200,0.6007226838005914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,65536,0.7628995577494303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,16384,0.20073244306776258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,12288,0.1533084445529514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,12288,0.4694702360365126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,16384,0.6262462404039171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,10240,0.14123822583092585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,8192,0.11137333181169297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,10240,0.39490400420294863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,8192,0.3172310988108317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,7168,0.0980257789293925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,65536,2.6901412540011935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,51200,2.0482559204101562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,6144,0.08191021945741442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,7168,0.27964621120029026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,5120,0.06804711288876004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,4096,0.0603333314259847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,3584,0.04935288760397169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,6144,0.2408675617641873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,3072,0.04603111081653171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,5120,0.19947555330064562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,2560,0.03772799836264716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,3584,0.13534488942888048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,2048,0.03190933333502875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,3072,0.11680800384945339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,1536,0.025320000118679468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,2560,0.09916266467836167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,2048,0.08060889111624824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,1024,0.016196444630622864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,768,0.013357333011097379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,1536,0.06178577740987142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,512,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,1024,0.04532177911864387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,256,0.008370666868156856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,768,0.03840533230039809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,512,0.028802666399214003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,256,0.02239377796649933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,128,0.00831200016869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,64,0.007286222444640265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,128,0.020304888486862183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,3072,32,0.007644444704055786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,4096,0.15614933437771267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,32,0.020213334096802604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,3072,64,0.020554666717847187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,65536,0.6314489046732584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,51200,0.5089164310031468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,16384,0.16893511348300508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,12288,0.1301742262310452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,16384,0.6252337561713325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,10240,0.11664533615112305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,12288,0.4509279992845323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,8192,0.09552355607350667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,10240,0.3737493356068929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,51200,1.9036649068196614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,8192,0.301512877146403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,7168,0.08082399765650432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,6144,0.07803999715381198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,5120,0.05880000193913778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,65536,2.4188249376085067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,6144,0.22834844059414336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,4096,0.04656799965434604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,7168,0.26553866598341197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,5120,0.18873155117034912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,3584,0.042974223693211876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,3072,0.03961155480808682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,4096,0.1478159957461887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,2560,0.03096888793839349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,3584,0.12828266620635986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,2048,0.026691555976867676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,3072,0.11149688561757405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,2560,0.09338666995366414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,1536,0.022294221652878657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,2048,0.0758204460144043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,1024,0.015860444969601102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,1536,0.05819733275307549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,768,0.013550221920013428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,1024,0.04354844490687052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,512,0.0109120003051228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,256,0.008354666332403818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,512,0.02725244396262699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,128,0.006980444822046492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,256,0.02144355575243632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,64,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,128,0.019261333677503798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2560,32,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,64,0.018888889087571036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,32,0.018937778141763475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2560,768,0.034512887398401894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,65536,0.4927546713087294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,51200,0.38707910643683535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,16384,0.13515022065904406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,16384,0.5623173183865017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,12288,0.09879200326071845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,65536,2.2375049591064453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,10240,0.0817671087053087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,12288,0.42083022329542374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,8192,0.06692800256941053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,7168,0.06473510795169406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,10240,0.3527671231163873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,8192,0.28469332059224445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,6144,0.053195556004842125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,7168,0.2507137722439236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,5120,0.043841779232025146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,4096,0.036178665028678045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,6144,0.21553156110975477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,5120,0.17836533652411568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,3584,0.03174933460023668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,3072,0.028090665737787884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,4096,0.13877422279781765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,2560,0.025320000118679468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,51200,1.7484683990478516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,3584,0.12037155363294814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,3072,0.10329688919915093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,2048,0.021007999777793884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,1536,0.01627200014061398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,2560,0.08747377660539414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,1024,0.012568888564904531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,2048,0.07108889023462932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,768,0.010709332923094431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,512,0.008649778034951951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,1024,0.039091553952958845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,256,0.006676444576846228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,768,0.03253688746028476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,1536,0.05355111095640394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,512,0.025404444999165002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,128,0.005605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,64,0.0052177777720822235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,2048,32,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,256,0.019711999429596793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,64,0.01757866640885671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,32,0.017160000072585214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,2048,128,0.01829866733815935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,65536,0.3910355567932129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,51200,0.3030960030025906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,16384,0.1047439972559611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,12288,0.07840622133678861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,16384,0.524254216088189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,12288,0.39620621999104816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,10240,0.0735093355178833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,8192,0.05362400081422594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,10240,0.3327208889855279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,7168,0.04998666710323758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,8192,0.2682097752888997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,6144,0.04215822286076016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,51200,1.6305475234985352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,5120,0.038070221741994224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,7168,0.23616533809238008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,4096,0.030396444929970637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,6144,0.20318133301205107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,3584,0.028660443094041612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,5120,0.16810666190253365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,4096,0.13028621673583984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,3072,0.022288888692855835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,3584,0.11325866646236843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,2560,0.018668444620238412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,2048,0.015827554795477126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,3072,0.09721689091788398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,2560,0.0821582211388482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,1536,0.012772444221708508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,1024,0.009676444033781687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,2048,0.06810666455162896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,1536,0.052087999052471586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,1024,0.03677777780426873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,768,0.007992000215583378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,512,0.006724444528420766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,768,0.030604445272021826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,65536,2.074903064303928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,512,0.023749333288934495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,256,0.005973333285914526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,128,0.005314666777849197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,64,0.004910222358173794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,256,0.018564444449212816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,128,0.016302223006884258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1536,32,0.005027555757098728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,64,0.0158595558669832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1536,32,0.015837333268589444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,65536,0.314900451236301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,51200,0.25059466891818577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,16384,0.08406311273574829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,16384,0.4930657810635037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,12288,0.06442399819691975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,12288,0.37303911315070254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,10240,0.054676443338394165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,51200,1.5156817966037328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,65536,1.9336800045437281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,8192,0.042596442831887134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,7168,0.036004443963368736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,10240,0.31324977344936794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,6144,0.030771556827757094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,8192,0.2525564564598931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,7168,0.22217244572109648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,5120,0.025633777181307476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,4096,0.02123111155298021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,6144,0.19123289320203993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,5120,0.15847288237677679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,3584,0.018624888526068795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,4096,0.12351466549767388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,3072,0.016167999969588388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,2560,0.013824000126785703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,3584,0.10678311189015706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,2048,0.0118222220076455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,3072,0.09148177835676406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,1536,0.010144000252087912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,2560,0.07765866650475396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,1024,0.008021333151393467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,2048,0.06360355350706312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,768,0.007290666302045186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,1536,0.04770222306251526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,512,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,1024,0.03382755650414361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,256,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,768,0.02851466668976678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,512,0.02239111065864563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,128,0.004625777817434735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,256,0.016946666770511203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,128,0.014828445182906257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,32,0.0042053332759274375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,64,0.014740443891949125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,1024,32,0.014125333891974555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,65536,0.20541511641608345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,51200,0.16621866491105822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,1024,64,0.004209777961174647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,16384,0.05225244495603773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,16384,0.4743653403388129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,12288,0.041393776734670006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,51200,1.4581848780314128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,10240,0.03509689039654202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,65536,1.8595937093098958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,8192,0.02767466670937008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,10240,0.3018435637156169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,8192,0.24346666865878633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,7168,0.21387999587588838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,6144,0.021351110604074266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,12288,0.35902221997578937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,5120,0.017857778403494094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,6144,0.1841742197672526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,4096,0.01552977826860216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,7168,0.025243555506070454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,5120,0.15225066079033747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,3584,0.014161777165200977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,4096,0.11880000432332356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,3072,0.012413333687517377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,2560,0.010769777827792697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,3584,0.10241511132982041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,2048,0.009349333743254343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,3072,0.08772444725036621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,1536,0.008064000142945183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,2560,0.0745297802819146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,1024,0.006613333192136552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,2048,0.06037511428197225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,1536,0.04521244433191088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,768,0.00590133335855272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,512,0.005294222384691238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,1024,0.032433778047561646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,512,0.02161866592036353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,128,0.003912888881233004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,256,0.016206221448050607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,768,0.02772177755832672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,64,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,32,0.0041884444653987885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,128,0.014451555079884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,64,0.014222222897741528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,768,32,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,65536,0.19467377662658691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,51200,0.15149866210089788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,768,256,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,16384,0.05494488941298591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,16384,0.4625440173678928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,12288,0.042691555288102895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,65536,1.8090879652235243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,51200,1.4171057807074652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,12288,0.3503928979237874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,10240,0.03642311029964023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,8192,0.030026667647891577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,7168,0.026123555170165166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,8192,0.23787911732991537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,6144,0.023680888944202002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,7168,0.2089057763417562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,5120,0.01997244523631202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,6144,0.1792524390750461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,5120,0.14848355452219644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,4096,0.016513778103722464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,3584,0.01475911173555586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,4096,0.11604621675279404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,3584,0.09979289107852513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,3072,0.013168888787428537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,2560,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,3072,0.08525866932339138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,2560,0.0722035566965739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,2048,0.010053333308961656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,10240,0.29464443524678546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,1536,0.00830577810605367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,1024,0.006639110959238476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,2048,0.05839288896984524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,768,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,1024,0.03148622314135233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,1536,0.04422310988108317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,512,0.004987555659479565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,256,0.004266666869322459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,768,0.026138666603300307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,128,0.0037253333462609183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,64,0.0033964444365766314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,512,0.021011556188265484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,256,0.015561777684423657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,128,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,512,32,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,65536,0.1162453360027737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,32,0.012801777985360889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,512,64,0.013278221918476952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,51200,0.09143822060690986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,16384,0.038035555018319026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,12288,0.02979733215437995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,16384,0.4445804489983453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,12288,0.33641333050198025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,10240,0.021031111478805542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,8192,0.017419555120997958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,10240,0.2827155590057373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,8192,0.22810310787624785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,7168,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,6144,0.013857777747843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,51200,1.3592631022135417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,5120,0.012139555480745105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,7168,0.20077155696021187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,4096,0.010051555931568146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,65536,1.736572477552626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,6144,0.1727386713027954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,3584,0.009387555221716562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,3072,0.008655110994974772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,5120,0.1426559951570299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,4096,0.11088533533944024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,2560,0.007701333198282454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,2048,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,3584,0.09575377570258246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,1536,0.006262222097979651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,3072,0.08182311058044434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,2560,0.0687368909517924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,2048,0.05552533268928528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,768,0.004942222187916438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,1536,0.04091466797722711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,512,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,1024,0.029789331886503432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,256,0.0038071109188927543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,768,0.025493333737055462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,512,0.019920888874265883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,256,0.01481244464715322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,64,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,128,0.013065777719020844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,32,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,64,0.01239377756913503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,256,32,0.012406222522258759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,65536,0.0874933335516188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,51200,0.06978399886025323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,1024,0.005304000029961268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,16384,0.02959733208020528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,256,128,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,12288,0.021026665965716045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,16384,0.43558488951789004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,12288,0.32952266269259983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,10240,0.017630222770902846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,8192,0.012503110700183444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,51200,1.3301333321465385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,10240,0.27672266960144043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,65536,1.6975848939683702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,8192,0.22399555312262642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,6144,0.010075555907355415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,7168,0.19660888777838814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,5120,0.00924622184700436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,6144,0.16940000322129992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,5120,0.14002133740319148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,4096,0.008055999875068665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,3584,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,4096,0.1086497770415412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,3072,0.006686222222116258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,7168,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,2560,0.006325333482689328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,3584,0.09298400084177653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,2048,0.005375999957323074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,3072,0.0794186658329434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,1536,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,2560,0.06759377982881334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,1024,0.004264889078007804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,2048,0.05429155627886454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,1024,0.029483556747436523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,768,0.02494488822089301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,512,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,512,0.01961777773168352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,256,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,256,0.01460000044769711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,128,0.0032622222271230486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,1536,0.04083022144105699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,768,0.003948444293604957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,64,0.012161778079138862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,32,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,32,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,65536,0.08895644214418198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,51200,0.06659999820921156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,128,128,0.012421333127551608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,16384,0.026502221822738647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,16384,0.43483466572231716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,12288,0.014924445086055331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,12288,0.3290231227874756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,51200,1.3305688434176974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,128,64,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,10240,0.012813333008024426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,8192,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,65536,1.6996764077080622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,7168,0.010460444622569613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,6144,0.009697777529557547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,8192,0.22312355041503906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,10240,0.27662399080064565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,7168,0.19614666038089326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,4096,0.007675555845101674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,6144,0.1689280006620619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,3584,0.0069857777820693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,4096,0.10841600100199382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,5120,0.13956711027357313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,3072,0.006620444357395172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,3584,0.09318488836288452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,2560,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,2048,0.005211555709441503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,3072,0.07964177926381429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,2560,0.06711377700169881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,1536,0.0047004446387290955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,2048,0.05342488818698459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,1024,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,768,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,1536,0.04080533319049411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,1024,0.028932445579104956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,512,0.0033937777496046494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,768,0.0247662216424942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,5120,0.008632889224423302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,512,0.018911111685964797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,128,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,256,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,64,32,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,128,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,65536,0.08367288774914211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,64,0.011795555551846823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,64,32,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,51200,0.0672524438963996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,16384,0.02462933295302921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,12288,0.013103110922707452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,16384,0.4347795645395915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,12288,0.32877333958943683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,10240,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,8192,0.010337777435779572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,10240,0.27621155314975315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,7168,0.009685333404276106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,8192,0.2227475643157959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,6144,0.009015999734401703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,51200,1.3303431404961479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,7168,0.19626488950517443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,6144,0.16893688837687174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,5120,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,4096,0.009662222531106737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,3584,0.008962666822804345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,65536,1.6983999676174586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,3072,0.008623999853928884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,4096,0.10836178064346313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,3584,0.09336088763342963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,2560,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,3072,0.07960622178183661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,2048,0.0052471111218134565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,2560,0.067358222272661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,1536,0.004589333302444882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,2048,0.05343200100792778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,1024,0.00423555572827657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,1536,0.04024444354905023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,768,0.0038488888078265717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,1024,0.02940000096956889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,512,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,768,0.024375110864639282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,256,0.0033742222521040174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,512,0.019311111834314134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,5120,0.13983999358283147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,128,0.0031831111345026228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,64,0.03219822380277846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,256,0.0140657772620519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,128,0.012392889294359418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,64,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3841,32,32,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3841,32,32,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,16384,4.037255181206597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,51200,12.371120876736112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,16384,5.5405468410915795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,16384,4.129934946695964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,51200,18.760591295030384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,12288,3.05384890238444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,51200,13.816797892252604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,10240,2.513401879204644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,12288,3.805067698160807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,10240,3.0997625986735025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,12288,3.0443982018364797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,8192,2.0097777048746743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,8192,2.3902354770236545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,10240,2.520666758219401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,7168,1.7507173750135634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,8192,2.0393333435058594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,6144,1.4860418107774522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,7168,2.0796150631374783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,6144,1.7600666681925456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,7168,1.7909412384033203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,5120,1.1147875256008573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,5120,1.4766737620035808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,6144,1.5431938171386719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,4096,0.8784729109870063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,5120,1.289409743414985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,3584,0.8099377950032552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,4096,1.1642790900336373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,4096,1.0448373158772786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,3072,0.686757352617052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,3584,0.9232524236043295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,3072,0.8801253106858996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,3584,1.020208888583713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,2560,0.5952764617072211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,3072,0.8072586589389377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,2560,0.7475982242160373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,2048,0.5160577562120225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,2560,0.6882524490356445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,2048,0.6100444263882107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,1536,0.3935253355238173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,1024,0.2775217692057292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,2048,0.573280864291721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,1536,0.4821057849460178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,1536,0.45514843198988175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,768,0.19690577189127603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,1024,0.40151556332906085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,1024,0.34635199440850156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,768,0.28647645314534503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,512,0.15247466829088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,256,0.11432800028059219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,512,0.24231555726793078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,768,0.2868284384409587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,512,0.22969865798950195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,128,0.09369688563876682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,256,0.21573955482906768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,256,0.1805057790544298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,128,0.20468800597720674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,65536,128,0.1400675508711073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,32,0.08558489216698541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,64,0.20852000183529326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,65536,32,0.2138835589090983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,65536,64,0.08182399802737765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,51200,9.006204393174913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,65536,11.641549852159288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,65536,14.107660081651476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,65536,19.801284790039062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,16384,2.974402745564779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,16384,4.381120893690321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,51200,15.192379421657987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,51200,10.54897986518012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,12288,2.1267884572347007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,16384,3.3592809041341147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,12288,3.034134123060438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,10240,1.934780544704861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,12288,2.4835262298583984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,10240,2.4862694210476346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,8192,1.5904568566216364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,10240,2.0426222483317056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,8192,1.8909804026285808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,7168,1.3817555109659831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,8192,1.6346924040052626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,7168,1.6647191577487523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,6144,1.1600329081217449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,7168,1.4501893785264757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,6144,1.4106266233656142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,5120,0.9475022421942817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,6144,1.2508488761054146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,5120,1.181247075398763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,4096,0.7588844299316406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,5120,1.0520808961656358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,4096,0.930723508199056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,3584,0.6618355645073785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,4096,0.8509990904066297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,3584,0.8175173335605197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,3072,0.5843893157111274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,3584,0.754719999101427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,3072,0.7029129134284126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,2560,0.5092346403333876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,3072,0.6589333216349283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,2560,0.597308423784044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,2048,0.3919440110524495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,2048,0.49172268973456484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,2560,0.5600248972574869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,1536,0.27348266707526314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,2048,0.4676853285895453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,1024,0.19315466615888807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,1536,0.386262231402927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,1536,0.3711724546220567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,768,0.15138399600982666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,1024,0.29073866208394367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,512,0.11592533853318955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,768,0.23015732235378691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,1024,0.28257955445183647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,768,0.2343848811255561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,512,0.19353155295054117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,256,0.087281776799096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,512,0.18813155757056343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,128,0.07357066869735718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,256,0.17244799931844076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,256,0.1474319961335924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,64,0.061991108788384326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,128,0.16279644436306423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,51200,32,0.06474222077263726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,51200,128,0.11496799521976048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,64,0.16547289159562853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,51200,32,0.1699155569076538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,51200,3.1049067179361978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,65536,3.817913055419922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,65536,7.140460544162327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,65536,5.846968756781684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,16384,1.0064284006754558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,51200,5.6543761359320746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,16384,1.6452995936075847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,51200,4.568460252549913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,12288,0.7509439786275228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,16384,1.479647954305013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,12288,1.1680338117811415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,12288,1.1058453453911674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,10240,0.9661493301391602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,10240,0.6267111036512587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,8192,0.5228542221917046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,10240,0.9247271219889323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,7168,0.4591680102878147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,8192,0.7529484430948893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,7168,0.6622844272189671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,8192,0.7464213371276855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,6144,0.35750312275356716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,7168,0.6684942245483398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,6144,0.564036422305637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,5120,0.29468978775872123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,6144,0.5706239806281196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,4096,0.22802755567762586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,5120,0.4707173241509332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,5120,0.4808924463060167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,4096,0.37099554803636337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,3584,0.21031999588012695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,4096,0.38598132133483887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,3072,0.17354577117496064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,3584,0.32319821251763237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,3584,0.34024710125393337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,2560,0.15197422769334581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,3072,0.2973528967963325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,3072,0.27996622191535103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,2048,0.12462666299608018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,2560,0.25337423218621147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,1536,0.10056355264451768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,2048,0.2121573289235433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,2048,0.1962933275434706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,1536,0.15439644124772814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,1024,0.06682755549748738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,1536,0.16935377650790742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,768,0.05542222327656216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,1024,0.11622933546702068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,1024,0.1297244495815701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,768,0.1078524457083808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,768,0.0935715569390191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,512,0.04079822368092007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,2560,0.23818310101826987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,256,0.031361778577168785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,512,0.07750311162736681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,512,0.08723111285103692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,128,0.02773422168360816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,256,0.06779911120732625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,256,0.06665688753128052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,128,0.06208088662889269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,64,0.025133333272404138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,16384,128,0.05397422115008036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,16384,32,0.026554667287402686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,64,0.06275288926230536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,16384,32,0.06424888637330797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,65536,2.9392142825656467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,51200,2.2809075249565973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,65536,6.011323716905381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,65536,4.921754625108507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,16384,0.7241048812866211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,16384,1.3279813130696614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,51200,4.559893290201823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,12288,0.5746142069498698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,16384,1.2512808905707467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,10240,0.49030044343736434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,12288,0.9511849085489908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,12288,0.9356755150689019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,8192,0.40206223063998753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,10240,0.8572284380594889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,51200,3.8460748460557728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,10240,0.7835662100050184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,7168,0.3516844378577338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,8192,0.6210995780097114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,8192,0.6365182134840224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,7168,0.5442790985107422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,6144,0.2844524383544922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,7168,0.5594044261508518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,5120,0.23722667164272734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,6144,0.4650719960530599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,6144,0.48497687445746523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,4096,0.18789954980214438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,5120,0.40854310989379883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,5120,0.3886462317572699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,3584,0.17440177334679496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,4096,0.3046657774183485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,4096,0.32798488934834796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,3072,0.147544887330797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,3584,0.28787822193569607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,3072,0.22974933518303764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,3584,0.2658497757381863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,2560,0.12736088699764675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,3072,0.25043821334838867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,2048,0.09869244363572861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,2560,0.1961359977722168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,2560,0.21511288483937582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,2048,0.16101955042945013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,1536,0.07720977730221219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,2048,0.17926932705773246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,1024,0.0565777752134535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,1536,0.14348088370429143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,1024,0.09575288825564915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,768,0.04731377628114489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,1024,0.1097244421641032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,1536,0.12745066483815512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,512,0.03427733315361871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,768,0.07670666774113973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,768,0.09211111068725586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,256,0.025183111429214478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,512,0.07484800285763211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,128,0.021661332911915247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,512,0.06401689185036553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,256,0.05375022358364529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,256,0.05825155311160617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,64,0.02367466688156128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,12288,32,0.022282666630215112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,128,0.050309333536359996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,64,0.050385776493284434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,12288,128,0.04729155699412028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,12288,32,0.05148533317777845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,51200,1.848465813530816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,65536,2.2634942796495223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,65536,5.27071295844184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,65536,4.505791982014974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,16384,0.5994666417439779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,51200,4.009464687771268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,12288,0.4476622475518121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,51200,3.532080968221029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,16384,1.174274656507704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,12288,0.8421937624613444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,16384,1.152258661058214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,10240,0.41330843501620823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,12288,0.8632168769836426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,8192,0.3490000036027696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,10240,0.6961804495917426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,8192,0.5510808626810709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,10240,0.7229342460632324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,7168,0.2566995620727539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,6144,0.2230764495001899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,7168,0.4852835337320964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,7168,0.5157608985900879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,6144,0.41788532998826766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,5120,0.19772533575693765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,6144,0.4472968843248155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,5120,0.34482577111985946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,5120,0.3764915466308594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,4096,0.14889689286549887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,8192,0.5829155710008409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,3584,0.1415146721733941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,4096,0.2714017762078179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,4096,0.3000017801920573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,3072,0.11872533957163493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,3584,0.2640853457980686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,3584,0.2367475562625461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,3072,0.2041857772403293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,2560,0.09440710809495713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,3072,0.23055820994906953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,2048,0.07887199852201673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,2560,0.17401333649953207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,2560,0.19739645057254365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,1536,0.06206399864620633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,2048,0.14321777555677626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,2048,0.16442488299475774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,1024,0.0480915539794498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,1536,0.13143466578589544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,1536,0.11385956075456406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,1024,0.08556889163123237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,768,0.034677333301968045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,1024,0.10066399971644084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,512,0.026418666044871014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,768,0.06923911306593153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,768,0.08531200223498875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,512,0.05613778034845988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,256,0.020688888099458482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,512,0.06906577613618639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,128,0.01827022267712487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,256,0.047278222110536366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,256,0.054302222198910184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,128,0.044221334987216525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,10240,128,0.04366488920317756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,64,0.015627556376987033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,10240,32,0.015576000014940897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,64,0.044088890155156456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,10240,32,0.044454223579830594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,65536,2.131331549750434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,51200,1.6715670691596136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,65536,4.3826649983723955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,16384,0.542710198296441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,65536,4.106010860866971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,51200,3.534700393676758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,12288,0.41626845465766055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,16384,1.0169110827975802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,16384,1.0488772922092013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,51200,3.219606187608507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,12288,0.7671448919508191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,10240,0.34978310267130536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,12288,0.7869431177775065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,8192,0.28233422173394096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,10240,0.6084968778822157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,7168,0.24861245685153535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,10240,0.660046206580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,8192,0.5345599916246202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,7168,0.4467964702182346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,6144,0.20938666661580405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,7168,0.47194491492377383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,5120,0.17224621772766113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,6144,0.36564355426364475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,8192,0.482859558529324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,5120,0.3024817837609185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,4096,0.13005510965983072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,5120,0.344121774037679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,4096,0.2377706633673774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,3584,0.12027288807762994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,4096,0.2736133204566108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,3072,0.1031528910001119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,3584,0.2077937788433499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,3584,0.24136532677544487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,2560,0.08784888850318061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,3072,0.17936444282531738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,3072,0.20957333511776396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,6144,0.40954134199354386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,2048,0.07230577866236369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,2560,0.15323377980126276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,2560,0.17895021703508165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,1536,0.05470577875773112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,2048,0.1501555575264825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,2048,0.12694399886661106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,1536,0.09965688652462429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,1536,0.12030311425526936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,768,0.031179553932613794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,1024,0.09206222163306342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,1024,0.07547733518812391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,768,0.06084089146720039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,768,0.07772355609469943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,512,0.02159822152720557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,256,0.017862222260899015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,512,0.048750221729278564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,512,0.06319022178649902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,128,0.01517599986659156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,256,0.04959466722276476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,256,0.039444443252351545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,128,0.035748445325427584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,64,0.019527110788557265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,1024,0.03999377621544732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,8192,32,0.016535111599498324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,64,0.03465244505140517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,8192,32,0.034330666065216064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,8192,128,0.039312889178593956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,65536,1.7321484883626301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,51200,1.3737209108140733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,65536,4.058993869357638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,16384,0.45029158062405056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,65536,3.977142333984375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,16384,0.9375929302639432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,12288,0.32792356279161244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,51200,3.1975750393337674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,16384,1.0142018000284831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,12288,0.6865635448031955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,10240,0.29926400714450413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,12288,0.7621413336859809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,8192,0.24232623312208387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,51200,3.1169946458604603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,10240,0.5649573538038466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,10240,0.6399866739908854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,8192,0.45003286997477215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,7168,0.2037280003229777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,6144,0.18530844317542183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,8192,0.5160711076524522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,7168,0.39635199970669216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,7168,0.4552515347798665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,5120,0.14222221904330784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,6144,0.34245954619513613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,6144,0.3946933216518826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,4096,0.11278043852912055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,5120,0.2826888826158312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,5120,0.33199556668599445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,3584,0.09937244653701782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,4096,0.2207697762383355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,3072,0.08627200126647949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,3584,0.19195289081997344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,3584,0.23239289389716256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,4096,0.26458221011691624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,2560,0.07531466748979357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,3072,0.16698400179545084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,2048,0.06185955471462674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,3072,0.20215911335415312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,2560,0.17256000306871203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,1536,0.048400001393424145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,2048,0.11714933978186713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,2048,0.1446817715962728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,1024,0.036235554350747004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,1536,0.09278133180406357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,1536,0.11560622851053874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,768,0.02645688917901781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,1024,0.07031200329462688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,1024,0.08857333660125732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,512,0.02028888960679372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,768,0.056112892097897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,768,0.07467999723222521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,256,0.016424889365832012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,512,0.0446631113688151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,512,0.060453335444132485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,128,0.013423999978436364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,256,0.03537688983811273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,256,0.047269334395726524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,64,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,2560,0.14273599783579508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,128,0.03186310993300544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,7168,32,0.011798222031858234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,7168,128,0.03557688991228739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,64,0.03186933199564616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,7168,32,0.03119644522666931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,51200,1.34680355919732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,65536,1.6768230862087672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,65536,3.8339945475260415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,16384,0.41528977288140195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,65536,3.709089067247179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,51200,2.970623016357422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,12288,0.3151173326704237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,51200,2.905268351236979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,16384,0.8625840081108941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,10240,0.25521334012349445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,16384,0.947530640496148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,12288,0.6268026563856337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,8192,0.1963440047370063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,12288,0.714304871029324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,10240,0.5242604679531521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,10240,0.5987048678927952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,8192,0.41628532939487034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,8192,0.4858169025844998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,7168,0.36720977889166934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,7168,0.42751465903388125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,6144,0.15078132682376438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,5120,0.12459911240471734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,6144,0.31532088915507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,6144,0.37104177474975586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,4096,0.09978310929404365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,7168,0.17937866846720377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,5120,0.26066043641832143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,5120,0.3108373218112522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,3584,0.0860053300857544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,4096,0.2045217752456665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,4096,0.24785423278808594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,3584,0.17722045050726998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,3072,0.07348177830378215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,3584,0.21787556012471518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,2560,0.06515910890367296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,3072,0.15309154987335205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,3072,0.18867911232842338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,2048,0.05390666590796577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,2560,0.1621146731906467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,2048,0.10859022537867229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,2560,0.13217600186665854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,1536,0.040059556563695274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,2048,0.13566399945153132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,1024,0.031015111340416804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,1536,0.10831999778747559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,1536,0.08630577723185222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,768,0.02346666653951009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,1024,0.06422755453321669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,1024,0.08304978079266019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,512,0.018397332893477548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,768,0.07063555717468262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,768,0.0513937771320343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,512,0.040047112438413836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,256,0.0140666663646698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,512,0.05712266763051351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,128,0.013886222408877479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,256,0.03177600105603536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,256,0.04301599992646111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,128,0.029305779271655615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,64,0.02858933475282457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,32,0.01315111087428199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,6144,32,0.028251555230882432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,6144,128,0.033899555603663124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,65536,1.1373590893215604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,6144,64,0.013056000073750814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,51200,0.9962239795260959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,65536,3.4661617279052734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,16384,0.31919201215108234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,51200,2.6413076188829208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,65536,3.4531377156575522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,16384,0.7846550941467285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,12288,0.23502310117085776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,16384,0.8826675415039062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,51200,2.698814180162218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,10240,0.20434755749172637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,12288,0.5747875637478298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,8192,0.15362666712866888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,12288,0.668923536936442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,10240,0.47802045610215926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,8192,0.3829893271128337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,7168,0.13855377833048502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,10240,0.5597004360622829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,8192,0.4532017707824707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,7168,0.33771644698248965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,7168,0.39871909883287215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,6144,0.29019199477301705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,5120,0.09647911124759251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,6144,0.3460720115237766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,5120,0.23985422982109916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,5120,0.2889084551069471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,4096,0.1877004437976413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,4096,0.22975288497077095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,6144,0.12477689319186741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,3584,0.06561688582102458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,3072,0.057413332992129855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,3584,0.16315110524495444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,3584,0.20288533634609648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,3072,0.1410879956351386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,2560,0.05192088749673632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,3072,0.1752880016962687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,2048,0.04242222176657783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,2560,0.11994489034016927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,4096,0.08086311154895358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,2560,0.15084089173210993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,1536,0.033153778976864286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,2048,0.09996533393859863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,2048,0.1264408826828003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,1024,0.02237333357334137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,1536,0.10126133097542657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,768,0.019524445136388142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,1536,0.07907999886406793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,1024,0.058192001448737256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,512,0.014185777968830533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,768,0.04577155576811897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,768,0.06609244479073419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,256,0.011175110936164856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,512,0.05203377869394091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,256,0.03933066791958279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,256,0.028852442900339764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,128,0.010048888623714447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,128,0.032249778509140015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,64,0.008959111240175035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,128,0.026492445005310908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,5120,32,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,64,0.025613332788149517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,5120,1024,0.07767289214664035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,32,0.025759999950726826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,5120,512,0.03635911146799723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,65536,1.1019119686550565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,51200,0.8822444279988607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,65536,3.076196458604601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,16384,0.2813919915093316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,51200,2.430271996392144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,65536,3.31842401292589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,51200,2.5963075425889754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,16384,0.7104658020867242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,12288,0.21161954932742646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,12288,0.5230817794799805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,16384,0.8485831154717339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,12288,0.6406062444051107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,10240,0.1614124510023329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,8192,0.12615288628472224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,10240,0.43704978624979657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,8192,0.34934398863050675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,10240,0.5378604465060765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,7168,0.10747022098965114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,8192,0.43572621875339085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,7168,0.30834934446546763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,6144,0.10038222206963433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,7168,0.3836666742960612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,5120,0.08689866463343303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,6144,0.2653297848171658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,6144,0.3333244323730469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,5120,0.21907111008961996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,4096,0.06701688634024726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,5120,0.27839554680718315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,3584,0.05946399768193563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,4096,0.17158044709099662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,4096,0.221398221121894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,3584,0.14876888857947454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,3072,0.050126221444871694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,3584,0.19426311386956108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,2560,0.04307377669546339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,3072,0.168595552444458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,3072,0.12821243868933782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,2048,0.03572622272703383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,2560,0.14468355973561606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,1536,0.028159111738204956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,2048,0.0890408886803521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,2048,0.12108977635701497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,1024,0.020602666669421725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,1536,0.09696800178951687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,1536,0.07189689079920451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,1024,0.050865779320398964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,768,0.01739022301303016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,1024,0.07424355877770318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,512,0.014459555347760519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,768,0.04146044452985128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,768,0.06211466921700371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,256,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,512,0.032610668076409235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,512,0.051233778397242226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,128,0.009000889129108852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,2560,0.10872089200549656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,256,0.03818666603830125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,256,0.025823111335436504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,64,0.010968888799349466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,128,0.023345778385798138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,4096,32,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,4096,128,0.031203554736243352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,32,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,4096,64,0.02342133389578925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,65536,0.8581866688198514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,51200,0.6979120042588977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,65536,2.875584920247396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,51200,2.19452879163954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,16384,0.23929243617587617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,65536,3.1842212677001953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,51200,2.491299523247613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,16384,0.6600160068935818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,12288,0.1733093394173516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,16384,0.8146568934122721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,10240,0.15843733151753744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,12288,0.49274667104085285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,12288,0.6151644388834635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,10240,0.4133768876393636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,8192,0.1142782237794664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,10240,0.5175333552890354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,7168,0.1026479999224345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,8192,0.3324097792307536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,8192,0.41888801256815594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,6144,0.08524888753890991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,7168,0.2933564451005724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,5120,0.07325155867470635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,7168,0.37014489703708225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,6144,0.25220712025960285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,6144,0.31997156143188477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,4096,0.062421335114373096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,5120,0.20886488755544028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,3584,0.053116444084379405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,5120,0.26736267407735187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,4096,0.16344088978237575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,4096,0.21278399891323516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,3072,0.04571911030345493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,3584,0.14159022437201604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,2560,0.03905155592494541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,3584,0.18700977166493735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,3072,0.12236800458696152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,3072,0.16174044873979357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,2048,0.03161599900987413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,2560,0.10352800289789836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,2560,0.13956177234649658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,1536,0.02606755495071411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,2048,0.08481955528259277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,2048,0.11634222666422527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,1024,0.019951999187469482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,1536,0.06808888912200928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,1536,0.09326933489905463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,768,0.015394666128688388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,1024,0.048082666264639966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,1024,0.07123555739720662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,512,0.011927111281288994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,768,0.060328887568579785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,768,0.039748443497551814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,256,0.00905600024594201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,512,0.0478942228688134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,512,0.03086755673090617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,128,0.00832711077398724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,256,0.024082667297787134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,256,0.038146668010287814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,64,0.008997333546479544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,128,0.021672000487645466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3584,32,0.009301332963837517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,64,0.021262221866183813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3584,32,0.021686222818162706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,65536,0.8359297646416558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3584,128,0.030469334787792627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,51200,0.6479937765333388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,65536,2.6692223019070096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,16384,0.21494221687316895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,51200,2.0195759667290583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,65536,3.0521778530544705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,12288,0.162937773598565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,16384,0.6231519911024306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,16384,0.7811742358737521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,10240,0.1351191070344713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,12288,0.4686124589708116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,10240,0.39228089650472003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,51200,2.390459484524197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,8192,0.09983288579516941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,12288,0.5924382209777832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,10240,0.4970960087246365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,7168,0.09046577745013767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,8192,0.3160933388604058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,6144,0.07455199956893921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,8192,0.4019964536031087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,7168,0.27828976843092174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,5120,0.0614373352792528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,7168,0.3562222321828206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,6144,0.2394604418012831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,6144,0.30646933449639213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,4096,0.05047466688685947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,5120,0.1978764401541816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,3584,0.0469084448284573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,5120,0.256478230158488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,4096,0.15552088949415419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,4096,0.20395821995205352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,3072,0.03822311096721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,3584,0.13468533092074925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,2560,0.0328053335348765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,3584,0.17946488327450225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,3072,0.11631200048658584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,3072,0.15560355451371935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,2048,0.02924444609218174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,2560,0.09718044598897298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,1536,0.022796443767017786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,2048,0.07941777838601007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,2560,0.13343466652764216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,2048,0.1118862231572469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,1024,0.01607733302646213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,1536,0.06076355775197347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,768,0.012891555825869242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,1536,0.08919644355773926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,1024,0.04614488946066963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,512,0.010442666709423065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,1024,0.06908622052934435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,768,0.037074665228525795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,768,0.058119999037848584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,256,0.00833155545923445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,512,0.028612444798151653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,512,0.045569777488708496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,128,0.007642666498819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,256,0.022646221849653456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,256,0.03599911265903049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,64,0.006822222222884496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,3072,32,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,3072,128,0.029684444268544514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,128,0.020742222666740417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,64,0.019856888386938307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,3072,32,0.020259555843141343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,65536,0.6257840262518989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,51200,0.4606906572977702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,65536,2.386132346259223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,51200,1.9143750932481554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,16384,0.15083377891116673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,51200,2.2862701416015625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,65536,2.9229759640163846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,16384,0.5903715557522243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,12288,0.11644000477261013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,10240,0.10124088658226861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,16384,0.7501457532246908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,12288,0.44449779722425675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,12288,0.5678355428907607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,10240,0.3718560006883409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,8192,0.08600266774495442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,10240,0.4781013594733344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,7168,0.07672711213429768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,8192,0.2998053232828776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,8192,0.3865617646111383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,6144,0.0646595557530721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,7168,0.26464443736606175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,7168,0.3407031165228949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,5120,0.050613333781560264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,6144,0.22752621438768175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,6144,0.2961591084798177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,5120,0.18713333871629503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,4096,0.04282755653063456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,5120,0.24671289655897352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,3584,0.035623110002941556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,4096,0.14644622802734375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,4096,0.19523110654619005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,3584,0.12690222263336182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,3072,0.031410667631361223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,3584,0.17145333025190565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,2560,0.027088888817363318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,3072,0.1093191107114156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,2560,0.0922551088862949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,3072,0.14967289235856798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,2048,0.02201333310869005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,2560,0.12762488259209528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,1536,0.018527110417683918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,2048,0.10724622011184692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,2048,0.07570488585366143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,1536,0.05724622143639458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,1024,0.013844444519943662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,1536,0.08578399817148845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,768,0.010935111178292168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,1024,0.04237777656979031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,1024,0.0661697785059611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,512,0.00907911111911138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,768,0.03446311089727614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,768,0.055959999561309814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,256,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,512,0.04461955693033007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,512,0.027112000518374976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,256,0.021321778496106465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,128,0.006610666712125142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,256,0.03496711121665107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,64,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,128,0.019070222973823547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2560,32,0.006274666637182236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2560,128,0.028548445966508653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,64,0.018611555298169453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2560,32,0.018608000543382432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,65536,0.5523146523369683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,51200,0.4366462230682373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,65536,2.2220240698920355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,51200,1.7458729214138453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,16384,0.14193244775136313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,16384,0.5563128789265951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,51200,2.284080081515842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,12288,0.10800088776482476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,65536,2.9171449873182507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,16384,0.7476897769504123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,10240,0.10096000300513373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,12288,0.41992889510260684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,8192,0.07815911372502644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,10240,0.3517928918202718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,12288,0.5671208699544271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,10240,0.4758933385213216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,7168,0.06956266694598727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,8192,0.2838382191128201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,6144,0.06095555755827162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,8192,0.38499289088779026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,7168,0.2498799959818522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,5120,0.04893777767817179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,7168,0.34043465720282656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,6144,0.21457421779632568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,4096,0.04057066639264425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,6144,0.294086217880249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,5120,0.17756088574727377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,3584,0.03696889016363356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,5120,0.2455564339955648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,4096,0.1383262210422092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,4096,0.19403821892208525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,3072,0.031481779283947416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,3584,0.11997421582539876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,2560,0.02720622221628825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,3584,0.1717848910225762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,3072,0.14800355169508192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,2048,0.023000889354281958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,2560,0.08654844760894775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,2560,0.12719910674624973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,2048,0.07115377982457478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,2048,0.10611644718382095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,1536,0.05419199996524387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,1536,0.08514133426878188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,1024,0.039043555657068886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,1024,0.0629991094271342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,768,0.010800888968838586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,3072,0.10315110948350693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,768,0.031850665807724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,768,0.0548017786608802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,512,0.009001777403884465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,1536,0.017966222431924608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,512,0.025368889172871906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,256,0.007070221834712558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,1024,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,512,0.043190220991770424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,128,0.006599111275540457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,256,0.03350133366054959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,2048,128,0.02748888896571265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,64,0.006299555715587404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,128,0.01797155539194743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,2048,32,0.006684444430801604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,64,0.017258667283587985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,32,0.01761777698993683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,65536,0.42388176918029785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,2048,256,0.01995466649532318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,51200,0.323670228322347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,65536,2.0752319759792752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,51200,1.6251600053575304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,16384,0.10756622420416938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,16384,0.5233084360758463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,12288,0.082223998175727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,16384,0.7140737639533149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,12288,0.3957653310563829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,10240,0.06427822510401408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,12288,0.5421155293782552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,65536,2.7797893948025174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,10240,0.3321288956536187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,51200,2.179133309258355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,8192,0.05277155505286323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,10240,0.45579200320773655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,7168,0.0478115545378791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,8192,0.26780356301201713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,6144,0.038880000511805214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,7168,0.2359084553188748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,8192,0.3697440094417996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,7168,0.3257688946194119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,5120,0.0331857767370012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,6144,0.2023235559463501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,4096,0.02773244513405694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,6144,0.28231467141045463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,5120,0.16680622100830078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,3584,0.02411288850837284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,5120,0.23617066277398002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,4096,0.1309395498699612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,4096,0.18599288993411592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,3072,0.021888888544506494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,3584,0.11292533079783122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,2560,0.018972444865438674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,3072,0.09692177507612441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,3584,0.1638017760382758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,3072,0.14127822717030844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,2048,0.015536889433860779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,2560,0.08203911119037204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,1536,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,2560,0.12187555101182725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,2048,0.06738489203982882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,1024,0.009571555587980483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,2048,0.10247466961542766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,1536,0.05193422238032023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,1536,0.0810862249798245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,768,0.008135110967689091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,1024,0.03608355588383145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,512,0.00665244460105896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,1024,0.061187558703952365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,768,0.05110133356518216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,256,0.005631999837027655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,512,0.02372444503837162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,512,0.04138488901986016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,256,0.018245332770877414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,128,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,256,0.032256888018714056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,64,0.004847110973464118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,128,0.016186666157510545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1536,128,0.02643377747800615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1536,32,0.005211555709441503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,64,0.01555288831392924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,32,0.016183111402723525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,65536,0.2879902256859673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,51200,0.2400560114118788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1536,768,0.030206223328908283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,65536,1.9326835208468969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,16384,0.0765591131316291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,51200,1.5126266479492188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,16384,0.4924151102701823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,51200,2.076180352105035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,12288,0.05824355284372965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,16384,0.6825964185926648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,65536,2.6542790730794272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,10240,0.050056000550587974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,12288,0.37267377641465926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,8192,0.03946844405598111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,12288,0.5178577635023329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,10240,0.31327999962700737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,10240,0.43593777550591367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,7168,0.03402577837308248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,8192,0.25244800249735516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,6144,0.029750221305423315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,8192,0.3529680040147569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,7168,0.2222222222222222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,7168,0.3118320041232639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,6144,0.19070043828752306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,5120,0.02535733415020837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,6144,0.2700542343987359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,4096,0.02337155573897892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,5120,0.1573422220018175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,5120,0.2250888877444797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,3584,0.019313777486483257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,4096,0.12329867151048447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,4096,0.17760888735453287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,3072,0.01680444512102339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,3584,0.15602933035956487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,3584,0.10638755559921265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,3072,0.09142488903469509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,2560,0.015237333046065437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,3072,0.13556267155541316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,2048,0.013639999760521783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,2560,0.07726044125027128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,2560,0.11644889248741998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,1536,0.011178666518794166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,2048,0.06340889135996501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,2048,0.09735733270645142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,1024,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,1536,0.07768978012932672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,1536,0.048009776406817965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,1024,0.033759110503726535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,768,0.007986666427718269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,1024,0.05860177675882975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,512,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,768,0.050127112203174166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,512,0.022392888863881428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,256,0.005683555371231503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,512,0.039444443252351545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,256,0.016915554801623028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,128,0.004940444396601783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,256,0.03127466638882955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,128,0.014868444866604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,1024,128,0.02548444436656104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,64,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,1024,32,0.004552888787455029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,64,0.014478221535682678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,768,0.028616888655556574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,1024,32,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,65536,0.20752444532182482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,51200,0.16434666845533583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,65536,1.8584098815917969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,51200,1.4595680236816406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,16384,0.05232622226079305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,16384,0.4741751352945964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,12288,0.04053066505326165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,51200,2.076774173312717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,65536,2.65323723687066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,16384,0.681853347354465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,10240,0.03436888919936286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,10240,0.30116356743706596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,8192,0.027079110344250996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,12288,0.35935555564032656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,12288,0.5177946620517307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,10240,0.43532800674438477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,7168,0.025050666597154405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,8192,0.24335021442837187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,6144,0.0210479994614919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,7168,0.3109866778055827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,8192,0.35321511162651914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,5120,0.018560000591807894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,6144,0.18365600374009874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,6144,0.26910310321384007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,4096,0.015185778339703878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,5120,0.1522631115383572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,5120,0.22463554806179473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,3584,0.013749333719412485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,4096,0.11853688293033177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,4096,0.1782311068640815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,3072,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,3584,0.10256177849239773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,3584,0.15593777762518987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,2560,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,7168,0.21390844715966117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,3072,0.13520444764031306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,3072,0.08832089106241862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,2048,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,2560,0.07363288932376437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,2560,0.11563021606869167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,1536,0.008110221889283922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,2048,0.09706133604049683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,2048,0.06101600329081217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,1024,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,1536,0.04571733209821913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,1536,0.07665155331293742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,768,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,1024,0.05680444505479601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,1024,0.032927112446890935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,768,0.026755554808510676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,768,0.048319998714658946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,512,0.004934222333961063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,256,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,512,0.03876711262596978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,256,0.016460445192125108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,512,0.021644444929228887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,256,0.03086400032043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,128,0.014440889159838358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,64,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,64,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,32,0.0038693332009845306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,768,32,0.013769778112570444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,65536,0.1634017758899265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,768,128,0.0038693332009845306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,768,128,0.02472355630662706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,51200,0.13241067197587755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,65536,1.806876500447591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,51200,1.4160497453477647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,16384,0.045137776268853076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,16384,0.46281957626342773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,65536,2.5192559560139975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,12288,0.04060088925891452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,12288,0.3501635657416449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,16384,0.6493902206420898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,10240,0.03515911102294922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,51200,1.973722669813368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,12288,0.4925582143995497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,8192,0.0281333327293396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,8192,0.23741867807176378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,7168,0.024676443801985845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,8192,0.3349493344624837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,10240,0.4154764546288385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,7168,0.208369771639506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,7168,0.295775122112698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,6144,0.021820444199774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,5120,0.018008000320858426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,6144,0.17961511347028944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,10240,0.2941208945380317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,6144,0.25573955641852486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,4096,0.016241777274343703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,5120,0.14869066079457602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,3584,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,5120,0.2137413289811876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,4096,0.11604532930586074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,4096,0.1683866712782118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,3072,0.012418666647540199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,3584,0.14796445104810926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,3584,0.09987555609809028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,2560,0.011300444602966309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,3072,0.08552177747090657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,2048,0.009688888986905416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,2560,0.109497778945499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,2560,0.0723173353407118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,1536,0.008006222546100616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,2048,0.05860977702670627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,2048,0.0913155542479621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,1024,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,1536,0.07259022527270846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,1536,0.044171555174721606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,768,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,1024,0.03180977702140808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,1024,0.05416355530420939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,512,0.005199111170238919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,768,0.02622311148378584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,768,0.04585599899291992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,256,0.0041893331540955436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,512,0.021012443635198805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,3072,0.12776445017920599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,512,0.03743288914362589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,256,0.015254222684436373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,128,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,256,0.029048889875411987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,64,0.0038737778862317405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,512,32,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,512,128,0.02348088887002733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,128,0.013767999907334646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,64,0.013063111239009433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,65536,0.1069271100891961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,512,32,0.013086222112178802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,51200,0.08543199963039821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,51200,1.3579813639322917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,16384,0.03678755627738105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,65536,1.7356515460544164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,16384,0.44447024663289386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,12288,0.02519911030928294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,51200,1.9641751183403864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,65536,2.509779612223307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,12288,0.3365040090348985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,10240,0.019306666321224637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,12288,0.48957331975301105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,16384,0.6468070877922906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,8192,0.01551644504070282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,10240,0.28279111120435924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,7168,0.013827555709415011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,10240,0.4119439919789632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,8192,0.2286897765265571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,8192,0.3336631192101373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,6144,0.012412444584899478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,7168,0.20097155041164824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,5120,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,7168,0.2942799992031521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,6144,0.17258221573299834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,4096,0.009301332963837517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,6144,0.2548728783925374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,5120,0.14228622118631998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,3584,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,5120,0.21225599447886148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,4096,0.11121866438123916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,4096,0.16769066121843126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,3584,0.09532355599933201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,3072,0.008034666379292807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,3584,0.14667288462320963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,2560,0.007304888632562425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,3072,0.08148622512817383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,3072,0.1269973384009467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,2560,0.06845066944758098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,2048,0.006310222049554189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,2560,0.1090737779935201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,1536,0.005648889061477449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,2048,0.05557777484258016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,2048,0.09110489156511094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,1024,0.004978666702906291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,1536,0.04159733321931627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,1536,0.07173333565394084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,768,0.004212444441186057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,1024,0.05350844396485222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,1024,0.030062221818500098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,768,0.025087111526065405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,512,0.0042160000238153665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,768,0.04560977882809109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,256,0.00433955548538102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,512,0.0199617776605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,512,0.03711644477314419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,128,0.0038159998754660287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,256,0.014841778410805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,256,0.02881244487232632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,64,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,256,128,0.02304800020323859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,256,32,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,64,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,65536,0.0894906653298272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,32,0.012537777423858643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,256,128,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,51200,0.07276622454325359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,65536,1.6984258227878148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,16384,0.030285331938001845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,65536,2.5092720455593533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,51200,1.3308596081203885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,12288,0.02038666605949402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,16384,0.6457066535949707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,12288,0.3297964466942681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,10240,0.014630221658282809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,51200,1.963653352525499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,16384,0.43510667483011883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,8192,0.012025777664449481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,12288,0.48964442147148984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,10240,0.2767377694447835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,7168,0.01110577748881446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,8192,0.22302132182651094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,10240,0.41145867771572536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,6144,0.010001777774757808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,8192,0.33359466658698184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,7168,0.19662843810187447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,5120,0.008974221845467886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,7168,0.2934888998667399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,6144,0.1689608891805013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,6144,0.25459734598795575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,4096,0.007637333538797166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,5120,0.13975377877553305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,3584,0.006991110742092133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,5120,0.2115831110212538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,4096,0.10820177528593276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,4096,0.1666168901655409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,3072,0.006691555596060223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,3584,0.1461884445614285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,2560,0.005909333212508096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,3072,0.12641066975063747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,3584,0.09332444270451863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,2048,0.0053004444473319584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,2560,0.06679022312164307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,2560,0.1084631085395813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,1536,0.004885333279768626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,2048,0.05418666534953647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,2048,0.09077866872151692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,1024,0.004209777961174647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,1536,0.04008533226119147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,1024,0.029312888781229656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,1024,0.05330488748020596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,768,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,1536,0.07109155257542928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,512,0.0037973332736227247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,768,0.024455111887719896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,768,0.04507466819551256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,256,0.00350577798154619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,3072,0.07920444673962064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,512,0.03639111253950331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,512,0.01961955593691932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,256,0.028508444627126057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,256,0.01443822185198466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,64,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3840,128,128,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,128,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,32,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,64,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,65536,0.08121511009004381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,51200,0.06590311394797431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,128,32,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,128,128,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,16384,0.0273964438173506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,16384,0.4348328908284505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,51200,1.3303617901272244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,12288,0.015296889675988091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,65536,1.6967670652601454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,10240,0.013049778011110095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,8192,0.01202666676706738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,12288,0.32926400502522785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,10240,0.27614222632514107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,8192,0.2229048940870497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,6144,0.00979911122057173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,7168,0.011041777829329172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,5120,0.00868800034125646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,7168,0.1963040033976237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,4096,0.007630222373538547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,6144,0.16882310973273384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,3584,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,5120,0.1396426624721951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,3072,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,4096,0.1080568896399604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,2560,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,3584,0.0926355587111579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,2048,0.0053119998839166425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,3072,0.07966578006744385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,2560,0.06606755654017131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,1536,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,2048,0.05379022161165873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,1024,0.004211555752489301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,768,0.003956444561481476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,1536,0.03961955507596334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,1024,0.029968000120586816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,512,0.0034977777136696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,768,0.024715556038750544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,512,0.019327999817000497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,128,0.003047111133734385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,256,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,64,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,64,32,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,128,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,64,0.011782222323947482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,65536,0.0803662207391527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,51200,0.06437333424886067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,16384,0.026828444666332666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,16384,0.43412444326612687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,12288,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,64,32,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,51200,1.3298258251614041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,10240,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,8192,0.009362666971153682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,65536,1.6967368655734594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,12288,0.32870133717854816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,7168,0.009398221969604492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,6144,0.008633777499198914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,10240,0.2760426733228895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,5120,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,8192,0.22325155470106336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,7168,0.19614488548702666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,6144,0.16902400387658012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,3584,0.008632889224423302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,5120,0.1396053367190891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,4096,0.1079217791557312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,3072,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,2560,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,3584,0.09300000137752956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,2048,0.0052639999323421055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,3072,0.07974488867653741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,2560,0.06694489055209689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,1536,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,1024,0.0041946665280395085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,2048,0.05362666646639506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,768,0.0038737778862317405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,1536,0.040047112438413836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,1024,0.0292524430486891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,768,0.024455111887719896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,256,0.0031840000301599503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,512,0.01893422173129188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,128,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,256,0.014152000347773233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,64,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,128,0.012055111428101858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,32,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,64,0.01178311142656538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3840,32,32,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,512,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3840,32,4096,0.009529777699046666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,16384,3.496263927883572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,16384,5.2643411424424915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,12288,2.6046640608045792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,10240,2.4468036227756076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,12288,3.6327713860405813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,51200,10.910861545138888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,8192,2.006222195095486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,10240,2.9416116078694663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,8192,2.2620160844590926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,7168,1.766177813212077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,7168,1.982699500189887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,6144,1.4394284354315863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,5120,1.2179270850287545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,6144,1.6781004799736872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,4096,0.9311457739935981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,5120,1.3968995412190754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,4096,1.1062471601698134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,3584,0.8021333482530383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,51200,18.166086832682293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,3072,0.7117999924553765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,3584,0.9689359664916992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,3072,0.832266648610433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,2560,0.6073137919108073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,2048,0.48133330874972874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,1536,0.3775866561465793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,2560,0.7093911170959473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,1536,0.4622906578911676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,1024,0.25785599814520943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,2048,0.5775084495544434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,512,0.14273067315419516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,768,0.19087910652160645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,1024,0.3448328971862793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,256,0.10710666577021281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,768,0.29225511021084255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,128,0.09160355726877849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,512,0.23118755552503797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,64,0.07911022504170735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,256,0.20718044704861113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,65536,32,0.08406222528881496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,128,0.19554932912190756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,64,0.19867822859022352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,65536,32,0.2034106651941935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,51200,8.472648620605469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,65536,10.791998969184029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,16384,2.7886827256944446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,12288,2.061269336276584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,51200,14.294038560655382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,16384,4.16061528523763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,12288,2.8911031087239585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,10240,1.9573040008544922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,10240,2.3513687981499567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,8192,1.5540871090359156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,65536,18.59719000922309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,8192,1.8101849026150172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,7168,1.4186444812350805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,6144,1.1928159925672743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,7168,1.5845342212253148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,5120,0.9484515719943576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,6144,1.3453422122531469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,4096,0.7606604364183215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,3584,0.6623288790384928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,4096,0.8815733591715494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,5120,1.1196773317125108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,3072,0.595993783738878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,2560,0.4696764416164822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,3072,0.6657573382059733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,2048,0.38097066349453396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,2560,0.5675173335605198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,1536,0.2946728865305583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,1024,0.1982133388519287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,3584,0.7725271119011773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,768,0.1500640047921075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,1536,0.36740711000230575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,1024,0.2847359975179036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,512,0.11290844281514485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,768,0.21864267190297446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,256,0.08538489209281074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,512,0.18565332889556885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,128,0.07155733638339572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,2048,0.4630719820658366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,64,0.060360888640085854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,128,0.15610844559139675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,51200,32,0.06336266464657254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,64,0.15786755084991455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,32,0.16152000427246094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,51200,256,0.16497688823276097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,51200,2.982959959242079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,65536,3.8546808030870228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,16384,0.9894871181911893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,12288,0.7509564293755425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,16384,1.5597235361735027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,12288,1.103032006157769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,51200,5.377097659640842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,8192,0.49842045042249894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,65536,7.018710242377387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,10240,0.9099084006415473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,10240,0.6024364365471734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,7168,0.46209245257907444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,8192,0.712532467312283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,6144,0.3595484362708197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,7168,0.627221319410536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,5120,0.3098062144385444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,6144,0.5778177579243978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,4096,0.2455244329240587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,3584,0.21069867081112334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,4096,0.3490773306952582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,3072,0.17175288995107016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,5120,0.44536622365315753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,3584,0.30485600895351833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,2560,0.14519288804796007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,2048,0.1129866706000434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,3072,0.2640666696760389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,1536,0.08902488814459907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,2048,0.19345511330498588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,2560,0.22566845681932238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,1024,0.06941777467727661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,1536,0.14716888798607722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,768,0.052657776408725314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,512,0.041045334604051374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,768,0.0893697804874844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,1024,0.1108551091618008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,256,0.029883556895785864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,512,0.07425244649251302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,128,0.025774222281244066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,64,0.02241244415442149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,128,0.05968799855973986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,16384,32,0.023102222217453852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,256,0.06417955292595758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,64,0.06044622262318929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,16384,32,0.06133421924379137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,51200,2.0923360188802085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,65536,2.6630853017171225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,16384,0.7066808806525336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,12288,0.5296151373121474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,16384,1.252402623494466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,10240,0.504051579369439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,12288,0.8998497856987847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,8192,0.36048176553514266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,65536,5.639933268229167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,10240,0.7406097518073188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,7168,0.3203217718336317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,8192,0.5856897566053603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,7168,0.5147004657321507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,51200,4.380232916937934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,5120,0.20973155233595106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,4096,0.16918044620090059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,6144,0.4406577746073405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,3584,0.14891289340125188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,5120,0.3638586733076308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,4096,0.286680883831448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,3072,0.13388089338938394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,6144,0.24612800280253092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,3584,0.2505902184380425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,3072,0.21638311280144584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,2048,0.10143555535210504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,2560,0.1856622166103787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,2048,0.1523617770936754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,1024,0.052687111828062266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,1536,0.1210133367114597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,768,0.03966844413015578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,1024,0.0904924472173055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,512,0.03015822172164917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,768,0.07400088840060763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,256,0.023191110955344305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,512,0.06128266784879896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,2560,0.1202195617887709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,128,0.020280889338917203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,64,0.01739377776781718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,256,0.05203022228346931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,32,0.01811022228664822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,128,0.04832177691989475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,12288,1536,0.07601244582070245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,32,0.048960887723498873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,12288,64,0.04828800095452202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,51200,1.6457671059502497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,65536,2.192464828491211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,16384,0.5770026842753092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,12288,0.4288293255700006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,16384,1.0977884928385417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,10240,0.34627556800842285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,12288,0.8145670890808105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,8192,0.2890346580081516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,10240,0.6585271093580458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,7168,0.24494666523403594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,65536,4.81419203016493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,51200,3.794737074110243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,6144,0.22341778543260363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,7168,0.45785776774088544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,5120,0.1964684459898207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,6144,0.3925226794348822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,4096,0.14739021990034315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,5120,0.3239946630265978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,3584,0.13882488674587673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,8192,0.5221138000488281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,4096,0.25457512007819283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,3072,0.11230133639441596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,3584,0.2225653330485026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,2560,0.09420888953738743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,2048,0.07732178105248345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,1536,0.06109333038330078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,3072,0.19314666589101157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,1024,0.04259999924235874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,2048,0.13584444257948133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,2560,0.1653653383255005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,768,0.03378666771782769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,1536,0.10826666487587823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,512,0.025937777426507738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,256,0.019783111082182992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,512,0.0544488893614875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,128,0.01716977854569753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,768,0.06656266583336724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,256,0.04585599899291992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,64,0.014807111687130399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,128,0.041577776273091636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,10240,32,0.014816888504558139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,64,0.04167822334501478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,32,0.04045688774850633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,10240,1024,0.08107288678487141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,65536,1.8959492577446833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,51200,1.40137693617079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,16384,0.4899333318074544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,16384,0.9524443944295248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,12288,0.3583048979441325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,10240,0.3112800121307373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,65536,4.262230343288846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,12288,0.6878604359096951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,8192,0.25527199109395343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,7168,0.20229866769578722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,10240,0.6415306727091471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,8192,0.45592620637681747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,51200,3.2832169002956815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,7168,0.4016542169782851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,5120,0.1431866619322035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,4096,0.11388799879286025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,6144,0.34426045417785645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,3584,0.10842666361067031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,5120,0.2853075663248698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,3072,0.09548622369766235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,4096,0.2236257659064399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,3584,0.19486844539642334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,2560,0.07376088698705037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,3072,0.16968977451324463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,2048,0.0643813345167372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,1536,0.04719822274314033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,6144,0.17597422334882948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,2560,0.14474844932556152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,1024,0.03492710987726847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,2048,0.11919200420379639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,768,0.026778666509522334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,512,0.020967110991477966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,1024,0.071070220735338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,256,0.016488000750541687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,768,0.058168000645107694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,1536,0.09568533632490371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,128,0.014275555809338888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,512,0.04653333293067085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,64,0.01239733315176434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,8192,32,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,256,0.03759555684195624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,64,0.03361599975162082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,128,0.034140444464153714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,8192,32,0.03289244572321574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,51200,1.2362453672620985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,65536,1.5887911054823134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,16384,0.4095297654469808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,12288,0.3189573287963867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,16384,0.879434691535102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,10240,0.2615288893381755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,12288,0.6440835528903538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,51200,3.0432942708333335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,8192,0.2023306687672933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,10240,0.5314186414082845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,7168,0.19587111473083496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,8192,0.42373156547546387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,6144,0.1676293346616957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,7168,0.3741768995920817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,5120,0.12700266308254665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,65536,3.9310531616210938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,4096,0.11379199557834202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,6144,0.31967555152045357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,3584,0.09351111120647854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,5120,0.26526933246188694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,4096,0.20697777801089814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,3072,0.07913688818613689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,2560,0.0695271094640096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,3584,0.18061510721842447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,2048,0.05875911315282186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,3072,0.15762844350602892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,1536,0.042152000798119434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,1024,0.03003466791576809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,2048,0.11086666584014893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,2560,0.1351484457651774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,1536,0.08865422010421753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,768,0.024840889705551997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,512,0.018862222631772358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,1024,0.0657786660724216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,256,0.014815110299322339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,128,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,512,0.04201777776082357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,64,0.015132443772421943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,256,0.034492443005243935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,7168,32,0.015517334143320719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,128,0.030590222941504583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,64,0.03020711077584161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,32,0.030204445123672485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,7168,768,0.053188443183898926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,51200,1.1253431108262804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,65536,1.4547351201375325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,16384,0.391391118367513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,12288,0.3058248890770806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,16384,0.8049386872185601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,10240,0.2561359935336643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,12288,0.5904240078396267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,10240,0.49127912521362305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,8192,0.1999911069869995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,65536,3.5847591824001737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,7168,0.16714133156670463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,51200,2.7672044965955944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,8192,0.3934657838609483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,7168,0.3454222149319119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,5120,0.1181866725285848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,6144,0.29731376965840656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,4096,0.09503644704818726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,5120,0.24399643474155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,3584,0.08349511358473037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,4096,0.19161066744062635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,3072,0.0747564435005188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,6144,0.14040977425045437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,3584,0.16675999429490831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,2560,0.06191644403669569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,3072,0.14486755265129939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,2048,0.051310221354166664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,1536,0.04119022356139289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,2560,0.12450133429633246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,1024,0.025990222891171772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,2048,0.10302044285668267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,768,0.02160888910293579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,1024,0.06026666694217258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,1536,0.08224622408548991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,512,0.01611911091539595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,768,0.04876177840762668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,256,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,128,0.013425778183672162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,256,0.03109244505564372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,512,0.038712001509136625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,128,0.027796443965699937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,32,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,64,0.027907556957668726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,6144,32,0.027526222997241553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,6144,64,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,51200,0.8647244241502551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,65536,1.0430640114678276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,16384,0.31241244739956325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,12288,0.2117208904690213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,16384,0.7351502312554253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,12288,0.5405688815646702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,10240,0.17355733447604707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,8192,0.144359999232822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,10240,0.45089954800075954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,7168,0.12236266665988499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,8192,0.36141244570414227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,6144,0.10954755544662476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,65536,3.2459182739257812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,7168,0.31755733489990234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,5120,0.09124000204934014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,51200,2.525930616590712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,4096,0.07580355803171794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,3584,0.06651200188530816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,6144,0.27353154288397896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,3072,0.06057955821355184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,4096,0.17553777164883086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,2560,0.04951911171277364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,3584,0.1534471114476522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,2048,0.042675554752349854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,3072,0.13299911552005345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,2560,0.11307999822828506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,1536,0.030996445152494643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,2048,0.0943688882721795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,1024,0.02259644369284312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,768,0.017850667238235474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,1536,0.07538844479454888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,512,0.014000889327791003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,1024,0.05424799852901035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,256,0.011071110765139261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,512,0.035106665558285184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,128,0.010004444254769219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,768,0.04598311252064175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,256,0.027448000179396734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,5120,0.22488623195224336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,64,0.0086986662613021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,5120,32,0.008650666309727563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,128,0.02537600033813053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,64,0.025029333101378545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,5120,32,0.025070221887694463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,51200,0.7626550992329916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,65536,0.9277475145128039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,16384,0.2615848912133111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,12288,0.19741866323682997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,16384,0.6623297797309028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,10240,0.14413511753082275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,12288,0.49236445956759983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,8192,0.12381154961056179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,10240,0.41063910060458714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,7168,0.10057422187593247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,51200,2.2461049821641708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,8192,0.32883556683858234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,6144,0.08933866686291164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,5120,0.07285688983069526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,7168,0.29037155045403373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,6144,0.24853687816196016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,4096,0.06270577510197957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,65536,2.917129728529188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,3584,0.052228444152408175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,4096,0.15950222810109457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,5120,0.20523377259572348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,3072,0.04651555418968201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,3584,0.1392266617880927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,2560,0.03862577676773071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,2048,0.031770666440327965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,2560,0.10232533348931207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,3072,0.12091644605000813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,1536,0.024633778466118708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,2048,0.08381688594818115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,1024,0.0177528891298506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,768,0.014578666951921252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,1024,0.048604445325003735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,512,0.011257777611414591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,1536,0.06842311223347981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,768,0.04007110993067423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,256,0.00945688866906696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,128,0.00811111099190182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,512,0.031424889961878456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,256,0.024749333659807842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,64,0.007278222176763747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,4096,32,0.007618666523032718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,64,0.022685334086418152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,128,0.02309155629740821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,4096,32,0.022039999564488728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,51200,0.6593155331081814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,65536,0.8195893499586316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,16384,0.219413333468967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,16384,0.6197439829508463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,12288,0.16900977823469374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,12288,0.464122666252984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,65536,2.632270177205404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,10240,0.14297244283888075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,51200,2.1190622117784286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,8192,0.11257155736287434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,7168,0.09932621982362534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,6144,0.08639644251929389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,10240,0.38998132281833225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,8192,0.313028441535102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,5120,0.07393511136372884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,4096,0.05864533450868395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,7168,0.2757599883609348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,6144,0.23621598879496256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,3584,0.054505778683556445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,5120,0.1946693393919203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,3072,0.04574399855401781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,2560,0.03857510950830247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,3584,0.1338248915142483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,2048,0.03253688746028476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,4096,0.15370933214823404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,1536,0.02574577762020959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,2560,0.09760444694095188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,1024,0.01973511113060845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,2048,0.07927822404437594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,768,0.015336889359686108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,1024,0.045979556110170156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,1536,0.0645457771089342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,768,0.03843466771973504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,256,0.008995555341243744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,512,0.029522667328516643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,256,0.023318222827381555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,128,0.007995555798212687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,64,0.007314666277832455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,128,0.021336888273557026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,32,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,3072,0.1158248848385281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,64,0.021341333786646526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3584,32,0.021223111285103693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3584,512,0.012136889000733694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,65536,0.6822977595859103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,51200,0.49473243289523655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,16384,0.18905333677927652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,12288,0.14451200432247585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,16384,0.5846906767951118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,10240,0.10913778013653225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,12288,0.4763760036892361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,8192,0.0891324414147271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,10240,0.3689280086093479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,7168,0.07687822315427992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,8192,0.2971377902560764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,6144,0.06676889128155179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,51200,1.9511369069417317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,7168,0.2617768976423475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,5120,0.0571920010778639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,65536,2.437136967976888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,4096,0.04622933268547058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,3584,0.040088001224729754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,5120,0.18471377425723604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,3072,0.03648622168434991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,4096,0.14580621984269884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,3584,0.12685155868530273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,2560,0.03151377704408433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,2048,0.025640888346566096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,3072,0.11003377702501084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,1536,0.020799999435742695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,2560,0.09245422151353623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,1024,0.01465777721669939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,2048,0.07496266894870333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,768,0.011571555501884885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,6144,0.22436266475253633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,768,0.03528889020284017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,512,0.009154666629102495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,512,0.027443556321991816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,256,0.008721777962313758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,256,0.02169955604606205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,128,0.0069297779765393995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,1024,0.04353422257635328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,64,0.006343110981914733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,128,0.01956622302532196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,3072,32,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,64,0.01960266629854838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,32,0.01961599952644772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,65536,0.5278488794962565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,3072,1536,0.05791111124886406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,51200,0.4134648905860053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,16384,0.15127110481262207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,16384,0.5728631019592285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,12288,0.10445688830481635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,12288,0.4173813396030002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,10240,0.09069244729148017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,51200,1.761998176574707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,8192,0.08178933461507161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,10240,0.3499457836151123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,7168,0.06612000200483534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,6144,0.05406044589148628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,7168,0.24797866079542372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,5120,0.047437333398395114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,65536,2.30622312757704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,6144,0.2136088874604967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,4096,0.03750133183267381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,3584,0.033371554480658636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,5120,0.17594756020439994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,3072,0.03121333320935567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,4096,0.13806666268242732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,2560,0.02514044443766276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,3584,0.12003466818067764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,2048,0.021201777789327834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,3072,0.10346133179134792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,1536,0.017093333933088515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,8192,0.28159912427266437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,1024,0.012396444049146442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,2560,0.08801866902245416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,2048,0.07015289200676812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,768,0.010701333483060202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,1536,0.053900443845325045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,512,0.008488000267081791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,256,0.0070186663005087115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,768,0.033569776349597506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,128,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,512,0.02625333269437154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,64,0.005672889037264719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,256,0.020744888318909537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,128,0.018625777628686693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2560,32,0.005954666684071223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,64,0.01827022267712487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,32,0.018328888548745047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,65536,0.45247199800279403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,51200,0.3591182231903076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,16384,0.13331377506256104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2560,1024,0.04087911049524943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,12288,0.09288533528645833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,16384,0.5377742449442545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,10240,0.07827999856736925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,12288,0.39334577984280056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,8192,0.061687999301486544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,10240,0.33056622081332737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,7168,0.05421688821580675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,8192,0.26658932367960614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,6144,0.045965333779652916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,51200,1.6363955603705511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,65536,2.0905492570665145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,5120,0.04152355591456095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,4096,0.03135555651452806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,7168,0.23460711373223198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,3584,0.02964977754486932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,6144,0.2007982201046414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,3072,0.025272889269722834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,5120,0.16606311003367105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,2560,0.02216622233390808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,4096,0.1302595535914103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,3584,0.11306311024559869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,2048,0.018365333477656048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,3072,0.09725244177712335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,1536,0.014682667122946845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,2560,0.08197066518995497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,1024,0.010919999745157031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,2048,0.0663715534740024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,768,0.009257777697510189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,1536,0.0524871117538876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,512,0.007648888561460707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,1024,0.03671911027696397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,256,0.006242666807439592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,128,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,512,0.0244186669588089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,64,0.004929777648713854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,256,0.01903555625014835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,2048,32,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,128,0.01756177842617035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,64,0.01720888912677765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,32,0.016897777716318767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,65536,0.3369617727067735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,51200,0.26208445760938853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,2048,768,0.031167109807332356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,16384,0.08851289086871678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,16384,0.4895342191060384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,12288,0.06948000192642212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,65536,1.9298852284749348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,10240,0.06801778078079224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,51200,1.5194568634033203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,12288,0.3698551124996609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,8192,0.05426311161783007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,7168,0.0463120010164049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,10240,0.3108675479888916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,8192,0.25079909960428876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,6144,0.043344888422224254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,5120,0.03535999854405721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,7168,0.22132978174421522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,4096,0.028152889675564233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,6144,0.1890817748175727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,3584,0.025309332542949255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,5120,0.15625066227383083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,3072,0.02314577831162347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,3584,0.10566488901774089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,2560,0.020023110840055678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,3072,0.09097599983215332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,2048,0.016713778177897137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,2560,0.07701866494284736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,1536,0.012833777401182385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,2048,0.06269066863589816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,1024,0.009485333330101436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,1536,0.048680888281928174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,768,0.007977777885066139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,1024,0.034345779154035784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,512,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,768,0.029179556502236262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,256,0.005659555395444234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,512,0.02295111119747162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,128,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,4096,0.12240978082021077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,256,0.018014222383499146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,64,0.00462400002612008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1536,32,0.0046364445653226645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,128,0.015949333707491558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,64,0.015455999308162265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,65536,0.24101334147983125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1536,32,0.015568888849682279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,51200,0.18795466423034668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,16384,0.06307822465896606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,12288,0.052538666460249156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,16384,0.45962222417195636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,10240,0.039656887451807656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,12288,0.34637244542439777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,8192,0.03247111042340597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,10240,0.2912151018778483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,8192,0.2354853418138292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,7168,0.02845066785812378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,6144,0.024925332930352952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,51200,1.411653306749132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,7168,0.20647733741336396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,6144,0.17720177438524035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,5120,0.021719111336602107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,65536,1.794832017686632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,4096,0.01759999990463257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,3584,0.01641688909795549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,3072,0.014378666877746582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,5120,0.14608355363210043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,3584,0.09881422254774307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,2560,0.012302222351233164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,4096,0.11397155125935872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,2048,0.010461333725187512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,3072,0.08439289198981391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,1536,0.00907199995385276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,1024,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,2560,0.0718782212999132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,768,0.006603555546866522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,2048,0.05890311135186089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,512,0.005623999983072281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,1024,0.03186044428083632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,256,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,768,0.027115555273161993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,128,0.0041893331540955436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,512,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,256,0.016756445169448853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,64,0.004179555508825514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,128,0.01479822231663598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,1024,32,0.003939555750952827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,64,0.014461333552996317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,32,0.01408088869518704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,65536,0.19076800346374512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,51200,0.15239289071824816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,16384,0.05400000015894572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,1024,1536,0.0444160004456838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,12288,0.03971111112170749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,16384,0.44597779379950625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,10240,0.03380444314744737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,12288,0.3382871150970459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,8192,0.02511377798186408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,10240,0.28385334544711643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,7168,0.022425777382320825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,51200,1.3698879877726238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,8192,0.22901866171095106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,6144,0.019748444358507793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,65536,1.7450035942925348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,5120,0.016350222958458793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,6144,0.1729226642184787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,7168,0.20163733429378936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,4096,0.015056000815497505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,5120,0.14226667086283365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,3584,0.012507555385430654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,3072,0.011506666739781698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,3584,0.09584533505969578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,4096,0.11088889174991184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,3072,0.082260443104638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,2048,0.008743999732865227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,2560,0.070470220512814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,2048,0.056194669670528836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,1536,0.00792711145348019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,1024,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,1536,0.042358222934934825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,768,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,1024,0.03119644522666931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,512,0.004879111217127906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,2560,0.010040000081062317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,256,0.0041955556306574075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,512,0.021014221840434607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,128,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,256,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,64,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,128,0.014164444473054675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,768,32,0.003949333396222857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,64,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,65536,0.15642933050791422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,32,0.013835555149449242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,768,768,0.02625511089960734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,51200,0.1263857815000746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,16384,0.04181155562400818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,16384,0.4276248878902859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,12288,0.03125244379043579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,51200,1.3063004811604817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,10240,0.025293333662880793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,65536,1.6682782702975805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,8192,0.019202666150199044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,7168,0.016993777619467843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,10240,0.271762662463718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,8192,0.2201777829064263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,6144,0.014670222997665405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,5120,0.012900444368521372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,7168,0.19326933224995932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,4096,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,6144,0.16591289308336046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,3584,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,5120,0.13648533821105957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,3072,0.008978666530715095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,4096,0.10579644309149848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,2560,0.00795911086930169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,12288,0.32426312234666615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,3584,0.091575112607744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,3072,0.07862844732072619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,1536,0.006326222171386083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,2560,0.06689688894483778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,2048,0.05336799886491564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,1024,0.0052382221652401825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,768,0.004599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,1024,0.029487109846538965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,512,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,768,0.025068443682458665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,256,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,512,0.020232000284724765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,128,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,256,0.015209777487648858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,64,0.0032586666444937387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,128,0.01349244432316886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,32,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,64,0.012859555582205454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,512,2048,0.007307555940416124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,65536,0.11282489034864639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,32,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,51200,0.09026933378643459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,16384,0.03756977783309089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,512,1536,0.04012444284227159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,12288,0.02769777841038174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,16384,0.41757867071363663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,12288,0.31624799304538304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,10240,0.022992889086405437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,8192,0.019484443797005545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,10240,0.26534665955437553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,7168,0.016574222180578444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,8192,0.21435111098819307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,6144,0.015125332607163323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,51200,1.2729084226820204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,5120,0.012725333372751871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,7168,0.18849778175354004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,65536,1.6262231402926977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,4096,0.00867022242810991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,3584,0.007972444097201029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,5120,0.13238310813903809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,3072,0.007275555696752336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,3584,0.08914044168260361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,2560,0.006616000086069107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,3072,0.0764791104528639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,2560,0.06509155697292753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,2048,0.005922666854328579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,2048,0.051670223474502563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,1536,0.0052435555391841466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,6144,0.1614675521850586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,1024,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,1536,0.039677331844965615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,768,0.00425155543618732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,4096,0.10330400202009414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,512,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,1024,0.028375999795065984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,768,0.024668445189793903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,256,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,128,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,512,0.01960444450378418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,256,0.014833778142929077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,64,0.003031999907559819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,128,0.01276711126168569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,64,0.012518222133318583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,65536,0.07937510808308919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,256,32,0.012146666646003723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,51200,0.06416710880067614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,256,32,0.0032026666320032547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,16384,0.02719377809100681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,51200,1.2442604700724285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,12288,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,16384,0.4078240129682753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,10240,0.014734221829308404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,12288,0.30857422616746694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,10240,0.2594710985819499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,8192,0.012511110968059964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,7168,0.011163555913501315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,8192,0.20978844165802002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,6144,0.009822222093741098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,7168,0.18384799692365858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,65536,1.5881200366550023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,4096,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,6144,0.15810044606526694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,3584,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,5120,0.1296044455634223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,4096,0.10018933481640285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,3072,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,2560,0.005961777849329843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,3584,0.08688800202475654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,3072,0.07446489069196913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,2048,0.005573333137565189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,1536,0.004936888813972473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,2048,0.0514355566766527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,2560,0.06311377551820543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,1024,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,768,0.003943110919660992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,1536,0.03911555475658841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,512,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,5120,0.009118222528033787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,256,0.00351555562681622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,1024,0.027958220905727808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,768,0.024119110571013555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,128,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,64,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,512,0.018957333432303537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,256,0.013929777675204806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,128,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,65536,0.0803404450416565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,32,0.011824000212881299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,128,64,0.01182577759027481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,51200,0.06242933538224962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,16384,0.025080889463424683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,16384,0.4071022139655219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,12288,0.013744889034165276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,128,32,0.0032595555401510666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,10240,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,12288,0.30810756153530544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,8192,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,51200,1.2437671025594075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,7168,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,65536,1.5872444576687281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,10240,0.25923199123806423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,6144,0.010060444474220276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,5120,0.008391111261314815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,7168,0.18378755781385633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,4096,0.007687999970383114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,6144,0.1578808890448676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,5120,0.1291537814670139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,3584,0.00701155596309238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,3072,0.006622222148709827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,3584,0.08665244446860419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,4096,0.09992177618874444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,2560,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,2048,0.005245333330498801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,3072,0.07454311185412936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,1536,0.004657777647177379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,2048,0.050287998384899564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,2560,0.06276000208324857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,1024,0.0038559999730851916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,768,0.0038622220357259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,1536,0.0387386679649353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,8192,0.2092506620619032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,512,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,768,0.023835556374655828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,256,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,1024,0.027909331851535376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,512,0.018590221802393597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,128,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,64,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,256,0.01423911088042789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,128,0.012157333393891653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,64,32,0.0033084443873829315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,64,0.011694221860832639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,65536,0.07744977871576945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,64,32,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,51200,0.06168711185455322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,16384,0.024295111497243244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,16384,0.40729689598083496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,12288,0.013452444639470843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,51200,1.2442862192789714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,10240,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,12288,0.30861245261298287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,8192,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,65536,1.5883973439534504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,7168,0.009329777624871995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,10240,0.2590364350212945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,6144,0.00869777798652649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,8192,0.20938311682807076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,5120,0.01828799976242913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,7168,0.18391733699374727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,6144,0.15796977943844265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,4096,0.010142222046852112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,3584,0.00902844468752543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,5120,0.1293075614505344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,4096,0.09988977511723836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,3072,0.00797333319981893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,2560,0.006953777538405524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,3584,0.08704533179601033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,2048,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,3072,0.07420800129572551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,1536,0.004884444591071871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,2560,0.06361955404281616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,1024,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,1536,0.03819466630617777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,2048,0.05003022154172262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,768,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,512,0.0033039999090962936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,1024,0.027950220637851294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,256,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,512,0.018592889110247295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,128,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,256,0.014104000396198697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,64,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,128,0.01200888885392083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3585,32,32,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,64,0.01181244436237547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,32,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3585,32,768,0.023784889115227595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,16384,3.6924658881293406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,16384,5.128273010253906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,51200,11.257738749186197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,12288,2.7505785624186196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,16384,3.8470514085557728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,51200,12.555154588487413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,51200,17.711161295572918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,12288,3.572729746500651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,12288,2.8636550903320312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,10240,2.289414299858941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,10240,2.891630172729492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,10240,2.3981270260281033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,8192,2.1255582173665366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,8192,2.2156079610188804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,7168,1.572013325161404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,8192,1.890933354695638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,7168,1.9423111809624567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,6144,1.3380088806152344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,7168,1.6889555189344618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,6144,1.6428613662719727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,5120,1.1263582441541884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,6144,1.4350257449679906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,5120,1.3769333097669814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,4096,0.8915315204196506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,4096,1.0854995515611436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,5120,1.2030275132921007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,3584,0.7663795683119031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,4096,0.975410673353407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,3584,0.9518124262491862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,3072,0.6833599938286675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,3584,0.8639413515726725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,3072,0.8244586520724826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,2560,0.5590728653801812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,2560,0.7062782181633843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,3072,0.756897767384847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,2048,0.4662693341573079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,2560,0.6428746647304958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,1536,0.36249245537651914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,2048,0.5714800092909071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,2048,0.5361119906107584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,1024,0.23195023006863066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,1536,0.4500560230678982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,768,0.18963644239637586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,1536,0.42795289887322324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,1024,0.3398257891337077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,1024,0.33407023217942977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,512,0.14990222454071045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,768,0.2692897849612766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,768,0.2682453261481391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,256,0.10668800274531047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,512,0.22868710094028047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,512,0.21569866604275176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,128,0.08743200037214492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,256,0.16875910758972168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,256,0.20450045002831352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,128,0.19300356176164415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,64,0.07507999738057454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,65536,128,0.13162044684092203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,65536,32,0.07734400033950806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,64,0.1956168942981296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,65536,32,0.19976177480485705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,65536,11.441121419270834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,51200,8.75317128499349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,65536,17.601086086697048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,65536,13.083147684733072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,16384,2.8093359205457897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,16384,4.083262125651042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,12288,2.240488052368164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,16384,3.104468451605903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,51200,13.964730156792534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,51200,10.07409413655599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,10240,1.8101529015435114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,12288,2.819935056898329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,12288,2.297472847832574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,10240,2.3087565104166665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,8192,1.404335127936469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,8192,1.7689822514851887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,10240,1.9097732967800563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,7168,1.1946461995442708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,8192,1.5334693060980902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,6144,1.0989724265204537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,7168,1.557109302944607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,7168,1.3548800150553386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,6144,1.3179733488294814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,5120,0.8659502135382758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,4096,0.6829013294643826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,6144,1.173160023159451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,5120,1.1056835386488173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,5120,0.9811422559950086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,3584,0.5872355567084419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,4096,0.8717448976304797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,4096,0.7966284222073026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,3584,0.7682524257236056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,3072,0.504433790842692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,3584,0.7047333187527127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,2560,0.4139751063452826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,3072,0.6702942318386502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,3072,0.6165786849127876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,2048,0.35968889130486387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,2560,0.5606106652153863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,2048,0.45790576934814453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,2560,0.5232675340440538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,1536,0.2550568845536974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,2048,0.4715840021769206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,1024,0.1790168947643704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,1536,0.36164622836642796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,1536,0.34859644042121035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,768,0.1444737778769599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,1024,0.2651706536610921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,1024,0.2724168830447727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,512,0.11240977711147732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,768,0.22008977995978463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,768,0.2166293329662747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,256,0.0831422209739685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,512,0.18323644002278647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,256,0.16306132740444607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,256,0.13818666670057508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,128,0.07163911395602755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,512,0.17698044247097441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,64,0.06066311068005032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,128,0.15411999490525988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,51200,32,0.06343199809392293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,51200,128,0.10854133632447983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,32,0.1592293315463596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,51200,64,0.15593155225118002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,51200,2.767968919542101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,65536,3.4865716298421225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,65536,6.9225980970594625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,65536,5.395372602674697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,16384,0.8456391228569878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,51200,5.275224049886067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,16384,1.537594689263238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,51200,4.237814161512587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,12288,0.6799759864807129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,16384,1.373355547587077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,10240,0.5755697886149088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,12288,1.0931360456678603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,12288,1.0266346401638455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,8192,0.4236906634436713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,10240,0.8986053466796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,10240,0.8634559843275281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,7168,0.4059617784288194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,8192,0.7076960139804416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,8192,0.6970346768697103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,6144,0.3451057804955377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,7168,0.6210968759324815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,7168,0.6164257791307237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,6144,0.5282942454020182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,5120,0.2881386544969347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,6144,0.5961813396877712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,5120,0.43984444936116535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,4096,0.23569067319234213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,5120,0.44623645146687824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,3584,0.20463289154900444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,4096,0.34606220987108016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,4096,0.35904354519314235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,3584,0.3088257842593723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,3584,0.31613243950737846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,3072,0.26188622580634224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,2560,0.14135821660359701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,3072,0.2771182325151232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,2560,0.2236871189541287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,2560,0.2366586791144477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,2048,0.11624088552263047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,3072,0.17853066656324598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,2048,0.18348799811469185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,1536,0.09241155783335368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,2048,0.19816977447933623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,1024,0.06432800160513984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,1536,0.1452702283859253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,1536,0.15883289443122015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,768,0.056045333544413246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,1024,0.10948622226715088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,1024,0.12131555875142415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,512,0.039055999782350324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,768,0.08829600281185573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,768,0.10184711217880249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,256,0.029373334513770208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,512,0.0735973318417867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,512,0.08232800165812175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,128,0.025470222036043804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,256,0.063137776321835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,256,0.06440177890989515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,64,0.021811554829279583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,16384,128,0.05197244551446703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,128,0.0592106646961636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,16384,32,0.022117333279715642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,64,0.05957688887914022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,16384,32,0.061200890276167125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,51200,2.277862124972873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,65536,2.9149191114637585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,65536,5.583089616563584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,65536,4.606084611680773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,16384,0.7144204245673286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,51200,4.251675499810113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,16384,1.2374373541937935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,12288,0.5513813230726454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,51200,3.6155467563205295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,16384,1.1732844246758354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,10240,0.425634675555759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,12288,0.8865520159403483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,8192,0.3204844527774387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,12288,0.8793475362989637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,10240,0.7338488896687826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,10240,0.7379173172844781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,7168,0.31487645043267143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,8192,0.579344908396403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,8192,0.5988133218553331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,6144,0.23887199825710723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,7168,0.5078008969624838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,7168,0.5281990898980035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,5120,0.22438311576843262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,6144,0.436264885796441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,6144,0.4564533233642578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,4096,0.17298932870229086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,5120,0.3605982197655572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,5120,0.3817946645948622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,3584,0.14770221710205078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,4096,0.2840497758653429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,4096,0.30816178851657444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,3584,0.2479626602596707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,3584,0.27122312121921116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,2560,0.11315200063917373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,3072,0.21566666497124565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,3072,0.23670665423075357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,2048,0.09566311041514079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,2560,0.18367200427585176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,2560,0.20344355371263292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,1536,0.06793155272801717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,2048,0.1508515543407864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,2048,0.16944444179534912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,1024,0.04825866553518507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,3072,0.1252275572882758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,1536,0.12003821796841091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,1536,0.1360942257775201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,768,0.038104888465669416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,1024,0.08976888656616211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,512,0.030469334787792627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,1024,0.10440444284015232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,768,0.073125335905287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,768,0.088136891523997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,256,0.02277155551645491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,512,0.0710737771458096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,512,0.060418665409088135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,128,0.019907555646366544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,256,0.055436445607079395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,256,0.05167999863624573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,128,0.04761066701677111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,12288,128,0.045597334702809654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,64,0.01702222228050232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,12288,32,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,64,0.048264887597825795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,12288,32,0.04930577675501505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,65536,2.2206177181667752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,51200,1.7452622519599066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,65536,4.921081966824002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,65536,4.208923763699002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,16384,0.5777511066860622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,16384,1.0907146665785048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,16384,1.0738773345947266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,12288,0.4566062291463216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,51200,3.7386398315429688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,12288,0.7829084396362305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,10240,0.3861768775516086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,12288,0.8113155364990234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,51200,3.3036941952175565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,8192,0.30645423465304905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,10240,0.6498977873060439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,10240,0.6761600176493326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,7168,0.25854844517178005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,8192,0.5157724486456977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,8192,0.5747093094719781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,6144,0.22697599728902182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,7168,0.4532826741536458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,7168,0.4870808919270833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,5120,0.18909688790639242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,6144,0.38871110810173887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,6144,0.42411645253499347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,5120,0.32107199562920463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,5120,0.34913065698411727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,4096,0.14892800649007162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,3584,0.1322071154912313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,4096,0.2523111237419976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,4096,0.2819306585523817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,3072,0.11017600033018325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,3584,0.22137867079840767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,3584,0.24808621406555176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,2560,0.0921964446703593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,3072,0.19134933418697783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,2048,0.07649777995215522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,2560,0.16380533907148573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,2048,0.13470489448971218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,1536,0.05716355641682943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,2048,0.15494844648573133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,1536,0.10693955421447754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,1536,0.12454488542344834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,1024,0.04303200046221415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,1024,0.0803955528471205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,1024,0.09594044420454238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,768,0.03269777695337931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,3072,0.2163511117299398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,512,0.025946666797002155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,768,0.06561155451668634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,768,0.08074755801094903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,2560,0.18584267298380533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,256,0.019567110472255282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,512,0.05382666654056973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,512,0.06564355558819242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,128,0.01700088878472646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,256,0.05137333273887634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,256,0.0454346670044793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,64,0.01477688882086012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,10240,32,0.01481599940194024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,128,0.040904889504114784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,10240,128,0.0421377784676022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,32,0.04080177678002252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,10240,64,0.04064888755480448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,65536,1.7618320253160265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,51200,1.4974738226996527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,65536,4.244317372639974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,16384,0.448961787753635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,65536,3.8141449822319875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,51200,3.277155558268229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,12288,0.325455109278361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,51200,2.9860301547580295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,16384,0.949586656358507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,10240,0.28364798757765025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,16384,0.9739946789211698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,12288,0.6851235495673286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,8192,0.22390309969584146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,12288,0.7338799900478787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,10240,0.6104248894585503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,10240,0.6153013441297743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,8192,0.45180357827080625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,8192,0.49936797883775497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,7168,0.39831111166212296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,6144,0.1753377781973945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,7168,0.44138667318556046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,5120,0.15035111374325222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,6144,0.34154489305284286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,6144,0.3807244565751817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,4096,0.11595911449856228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,5120,0.2823653221130371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,7168,0.19879999425676134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,5120,0.31658487849765354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,3584,0.10838933123482598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,4096,0.2552311155531141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,4096,0.22187999884287515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,3584,0.19371378421783447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,3072,0.08619822396172418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,3584,0.22554487652248809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,3072,0.1678097777896457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,2560,0.07716621955235799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,3072,0.1972177823384603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,2048,0.06337422132492065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,2560,0.1437928941514757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,2560,0.16839822133382162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,1536,0.05009777678383721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,2048,0.14047822687360975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,2048,0.1189991103278266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,1024,0.03283910950024923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,1536,0.09417421950234307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,1536,0.11326489183637832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,768,0.025920000341203477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,1024,0.07100533114539252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,1024,0.08739644289016724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,512,0.02023555503951179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,768,0.05750399827957153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,768,0.07412799861696032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,256,0.016342222690582275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,512,0.05964710977342394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,256,0.04657155606481764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,128,0.014440000057220459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,256,0.0376835564772288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,128,0.03471555643611484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,64,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,8192,128,0.03915111223856608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,64,0.03288355469703674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,8192,32,0.012522666818565793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,32,0.03249511122703552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,8192,512,0.046465777688556247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,65536,1.7982497745090062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,51200,1.3437599605984156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,65536,3.7675370110405813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,16384,0.4370933373769124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,65536,3.6840968661838107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,51200,2.9769164191352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,12288,0.3276248772939046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,16384,0.8691368632846408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,16384,0.9398719999525281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,51200,2.886162651909722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,10240,0.2796604368421766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,12288,0.6351528697543675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,12288,0.7069359885321723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,8192,0.20058400101131865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,10240,0.5276017718844944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,8192,0.420527113808526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,10240,0.5959093305799696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,7168,0.18585067325168184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,6144,0.16394489341311985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,8192,0.48186397552490234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,7168,0.3703884548611111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,7168,0.4259457853105333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,5120,0.13636088371276855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,6144,0.3191146585676405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,4096,0.10859377516640557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,6144,0.36794577704535586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,5120,0.2625955475701226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,5120,0.304986662334866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,3584,0.08596799770991008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,4096,0.20498666498396131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,3072,0.07737600141101413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,3584,0.17873156070709229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,3584,0.21709956063164604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,4096,0.24631288316514757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,2560,0.06778044170803495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,3072,0.15643289354112414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,3072,0.18931733237372506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,2048,0.055854221185048424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,2560,0.16219466262393528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,2560,0.1345857779184977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,1536,0.042392889658610024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,2048,0.11054311196009318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,2048,0.13505244255065918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,1024,0.02951822347111172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,1536,0.10940977599885728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,1024,0.06568088796403673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,1024,0.08378755384021336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,768,0.024246222443050806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,512,0.019096889429622226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,1536,0.08822755681143866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,768,0.052433778842290245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,768,0.07062577539020114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,256,0.014448000325096978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,512,0.042020443412992686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,128,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,256,0.0332551101843516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,512,0.0580000016424391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,256,0.044293334086736046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,64,0.011872000164455838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,128,0.030602667066786025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,7168,32,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,7168,128,0.03509066502253214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,64,0.030625777112113103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,7168,32,0.029818667305840388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,51200,1.0726373460557725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,65536,1.4842177497016058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,65536,3.449821260240343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,65536,3.4234125349256725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,51200,2.7198577457004123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,16384,0.38198044564988876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,16384,0.8000115818447537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,12288,0.262664000193278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,16384,0.8753937615288628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,12288,0.5852470927768284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,10240,0.23701156510247123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,12288,0.6595262421502007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,51200,2.6771041022406687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,8192,0.19343111250135633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,10240,0.48758400811089414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,10240,0.5554764535692003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,8192,0.38894044028388125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,7168,0.16117778089311388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,8192,0.44954665501912433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,6144,0.13810843891567653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,7168,0.3427893320719401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,6144,0.29375465710957843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,5120,0.11531021859910752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,6144,0.3434444533454047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,4096,0.09409155448277791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,5120,0.24317065874735513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,5120,0.2863955497741699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,3584,0.08366311258739895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,4096,0.22860354847378203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,4096,0.19026488727993437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,3072,0.07171110974417792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,3584,0.20114755630493164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,3584,0.1656008826361762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,2560,0.05941510862774319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,7168,0.3960702154371474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,3072,0.14405332671271429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,2048,0.04869066675504049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,2560,0.12382311291164821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,2560,0.15055378278096518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,1536,0.038754665189319186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,2048,0.1027911106745402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,2048,0.12629244062635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,1536,0.08110844426684909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,1024,0.02702577743265364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,1536,0.10281066762076484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,768,0.022226666410764057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,1024,0.060328887568579785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,1024,0.07867733637491862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,512,0.01604622271325853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,768,0.06614310873879327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,512,0.037709332174725003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,768,0.0487022234333886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,256,0.012527999778588613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,3072,0.17555022239685059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,128,0.011006222003036074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,512,0.0537431107627021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,256,0.03966310951444838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,256,0.03031555480427212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,6144,128,0.03324977888001336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,64,0.009720000127951304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,128,0.027426666683620874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,6144,32,0.009722666607962714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,32,0.02715911136733161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,6144,64,0.027514666318893433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,65536,1.2301173739963107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,51200,0.8997849358452691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,65536,3.123183992173937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,16384,0.3005137708452013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,51200,2.495340347290039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,65536,3.287535137600369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,12288,0.22793067826165092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,16384,0.7274453375074598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,51200,2.5731048583984375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,10240,0.19228977627224395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,16384,0.8409404224819608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,12288,0.5372782283359104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,12288,0.6343822479248047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,8192,0.1608017815483941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,10240,0.44657156202528214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,10240,0.5324151251051161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,7168,0.13134577539232042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,8192,0.3583057721455892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,8192,0.4328800042470296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,6144,0.11010755432976617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,7168,0.315595547358195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,5120,0.09611111217074925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,6144,0.27116089397006565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,7168,0.3819137679206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,6144,0.32942401038275826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,4096,0.08176266484790377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,5120,0.22387290000915527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,3584,0.06453600194719103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,5120,0.27395110660129124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,4096,0.21987556086646187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,3584,0.15209955639309353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,3584,0.1938168870078193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,3072,0.0594995551639133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,2560,0.050182223320007324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,3072,0.13123377164204916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,3072,0.16865511735280356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,2048,0.04388622111744351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,2560,0.14417333073086208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,4096,0.17458400461408827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,2048,0.09435911311043634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,1536,0.0310791101720598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,2048,0.12108266353607178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,1536,0.07463466458850436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,1024,0.02258133391539256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,1536,0.09792088800006443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,1024,0.05458044343524509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,768,0.01797777745458815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,1024,0.07489599784215291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,2560,0.11275111304389106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,512,0.013657777673668332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,768,0.04327733318010966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,768,0.06316088967853122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,256,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,512,0.05106755428844028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,256,0.02753777801990509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,512,0.034607112407684326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,128,0.009701333112186855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,256,0.03788888785574172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,64,0.008658666577604082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,5120,128,0.03195022212134467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,5120,32,0.008743999732865227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,128,0.02550044490231408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,64,0.02474311159716712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,5120,32,0.02476355599032508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,65536,0.952480845981174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,51200,0.8183049096001519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,65536,2.881044387817383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,16384,0.24035734600490996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,65536,3.025310304429796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,16384,0.686789353688558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,51200,2.2462505764431424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,12288,0.1752328872680664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,16384,0.7776888741387261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,10240,0.15584177441067165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,12288,0.499342229631212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,51200,2.3833840688069663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,12288,0.5910897784762913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,8192,0.13048622343275282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,10240,0.41446577178107363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,8192,0.3268044524722629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,7168,0.10374222199122111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,10240,0.4933404392666287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,8192,0.3996933301289876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,6144,0.09152533610661824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,7168,0.2884595659044054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,5120,0.07189066542519464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,7168,0.35237601068284774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,6144,0.24772355291578504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,4096,0.06183289157019722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,6144,0.3050453397962782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,5120,0.25107910897996694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,4096,0.15914132859971789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,3584,0.051432000266181103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,3584,0.13875288433498806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,3584,0.17762577533721924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,4096,0.2028097841474745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,3072,0.0448017782635159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,2560,0.03803022371398078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,3072,0.1206631130642361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,3072,0.1564142174190945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,2560,0.10221866766611735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,2048,0.032872888776991106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,2560,0.13339911566840276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,1536,0.026177777184380427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,2048,0.08315733406278822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,5120,0.20341155264112684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,2048,0.11263467205895318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,1024,0.01797155539194743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,1536,0.06717422273423937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,768,0.014639111028777229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,1536,0.09107466538747151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,1024,0.0694248874982198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,512,0.01111822244193819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,768,0.03945599993069967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,1024,0.04905777838495043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,768,0.05880800220701429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,256,0.009053332938088311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,512,0.03088444471359253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,512,0.047212445073657565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,128,0.008004444340864817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,256,0.024718221690919664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,64,0.007302222152551015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,256,0.03682133224275377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,4096,128,0.03059200114674038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,4096,32,0.007695111135641734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,64,0.022270222504933674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,32,0.02204533252451155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,65536,0.9086880154079862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,4096,128,0.022634666826989915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,51200,0.7762319776746961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,65536,2.714391920301649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,16384,0.23442310757107207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,51200,2.0724879370795355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,65536,3.022942225138346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,12288,0.16494399971432155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,16384,0.7733173370361328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,12288,0.4651173485649957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,51200,2.366060468885634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,10240,0.14873600006103516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,12288,0.5844684706793891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,16384,0.6388897895812988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,8192,0.1122079955206977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,10240,0.38982754283481175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,10240,0.49161868625217015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,8192,0.3115653461880154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,7168,0.10059999757342869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,8192,0.39898043208652073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,7168,0.2750177648332384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,6144,0.08502399921417236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,7168,0.3512275483873155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,5120,0.07378488779067993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,6144,0.23609688546922472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,6144,0.30552977985805935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,5120,0.19354400369856092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,4096,0.05694578091303507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,5120,0.2520364390479194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,3584,0.05195555422041151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,4096,0.15174667040506998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,4096,0.2021199994617038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,3584,0.13217422697279188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,3584,0.17680888705783418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,3072,0.044455111026763916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,2560,0.03809955716133118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,3072,0.11418399545881484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,3072,0.15488889482286242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,2048,0.031686223215527005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,2560,0.1328808863957723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,2560,0.09755733278062607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,2048,0.07856800158818562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,1536,0.025275554921891954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,2048,0.11146310965220134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,1024,0.018668444620238412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,1536,0.06305955515967475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,1536,0.09096178081300522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,768,0.014663111832406787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,1024,0.045928888850741915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,1024,0.06836799780527751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,512,0.011184000306659274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,768,0.058218664593166776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,512,0.02972444560792711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,768,0.038219554556740656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,256,0.009992000129487779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,512,0.04531733194986979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,128,0.009671111073758867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,256,0.03517244590653314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,64,0.008971555365456475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,256,0.023370666636361018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,128,0.021368889345063105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3584,32,0.009692444569534725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3584,128,0.02972088919745551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,64,0.020947555700937908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3584,32,0.021002666817771062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,65536,0.7342124515109592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,51200,0.5383902125888401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,65536,2.479175991482205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,51200,1.8923083411322699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,16384,0.19536799854702422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,65536,2.888943142361111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,12288,0.13742578029632568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,16384,0.5837182468838161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,16384,0.7747422324286567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,10240,0.12342311276329888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,12288,0.4386800130208333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,12288,0.5619866583082411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,10240,0.3684319920010037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,8192,0.09625955422719319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,51200,2.2640177408854165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,10240,0.47145689858330625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,7168,0.09072089195251465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,8192,0.2961706585354275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,6144,0.06841422451866998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,7168,0.26023732291327584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,8192,0.38271021842956543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,5120,0.059750219186147056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,6144,0.2235271135965983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,6144,0.29083021481831867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,4096,0.04682488905058967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,5120,0.2408977879418267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,5120,0.1837537818484836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,3584,0.040200889110565186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,4096,0.14446489016215006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,4096,0.19294133451249865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,3072,0.03508977757559882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,3584,0.1696924500995212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,3584,0.12604355812072754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,2560,0.02938577863905165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,3072,0.10833511087629531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,7168,0.337752898534139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,3072,0.14859288268619114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,2048,0.025013332565625507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,2560,0.12737422519259983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,2560,0.09175644318262736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,1536,0.020678222179412842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,2048,0.07460266351699829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,2048,0.10688622130288018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,1024,0.014397333065668741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,1536,0.08586755726072524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,1024,0.04285866684383816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,1536,0.057849778069390186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,768,0.011969777444998423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,768,0.055789331595102944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,768,0.035389333963394165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,512,0.027494221925735474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,512,0.04389511214362251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,256,0.009141333401203156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,1024,0.06592533323499891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,256,0.03370133373472426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,128,0.0069546666410234236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,256,0.02200177808602651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,512,0.00941066692272822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,64,0.006325333482689328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,128,0.01994400057527754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,3072,32,0.006439111298984951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,3072,128,0.029010666741265193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,64,0.01960533360640208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,3072,32,0.019626667102177937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,65536,0.5615057945251465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,51200,0.44387290212843156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,65536,2.282123565673828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,16384,0.148744887775845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,51200,1.7556453280978734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,65536,2.7573714786105685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,16384,0.5525831116570367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,12288,0.11369778050316705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,12288,0.4161484506395128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,16384,0.7100835376315646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,10240,0.09469955497317845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,12288,0.5367209116617838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,8192,0.08642666869693333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,10240,0.3486337661743164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,10240,0.45115020540025497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,7168,0.06991822189754911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,51200,2.161410649617513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,8192,0.3664373291863336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,7168,0.24709865781995985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,6144,0.060954663488599986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,7168,0.3232231140136719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,5120,0.05115288827154371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,6144,0.21178666750590006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,6144,0.27930490175882977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,5120,0.17479288578033447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,4096,0.040941334433025785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,5120,0.23085776964823404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,3584,0.03588177760442098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,8192,0.2804204357994927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,4096,0.13648177517784968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,3072,0.031028442912631567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,4096,0.18528710471259224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,3584,0.11819467279646133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,3584,0.16242310735914442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,2560,0.02607288956642151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,3072,0.10236622227562799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,3072,0.14240177472432455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,2560,0.08684533172183567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,2560,0.1217999988132053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,1536,0.0177502218219969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,2048,0.10198933548397487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,1536,0.05382755398750305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,2048,0.07056444221072726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,1024,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,1536,0.08240799771414863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,768,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,1024,0.06272977590560913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,1024,0.04177777634726631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,768,0.03344444433848063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,512,0.008707555631796518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,768,0.05204533206091987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,256,0.006948444578382704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,512,0.025732444392310247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,512,0.04301599992646111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,256,0.020663999848895602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,2048,0.02241955531968011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,128,0.0063031112982167145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,256,0.032947556840048894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,128,0.018561777141359117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2560,128,0.027822222974565294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,32,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,64,0.017868444323539734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2560,32,0.01826400061448415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,65536,0.4362488852606879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,51200,0.3277253309885661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2560,64,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,65536,2.079210705227322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,16384,0.12722222010294595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,51200,1.6257058249579535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,16384,0.519669320848253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,65536,2.6314478980170355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,12288,0.09087110890282525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,16384,0.6753066910637749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,51200,2.061593797471788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,10240,0.08434755272335476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,12288,0.3929262161254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,8192,0.06389244397481282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,12288,0.5127235518561469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,10240,0.32893334494696724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,10240,0.4311066733466254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,7168,0.05460977885458204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,8192,0.2656444443596734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,8192,0.3497182263268365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,6144,0.04920711119969686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,7168,0.23306488990783691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,7168,0.3088017834557427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,5120,0.042071110672420926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,6144,0.19962222046322295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,6144,0.26633421579996747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,4096,0.03158311049143473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,5120,0.16481510798136392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,5120,0.21986577245924208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,3584,0.028981334633297388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,4096,0.1285626623365614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,4096,0.17567111386193168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,3072,0.02474044428931342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,3584,0.11159022649129231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,3584,0.15510577625698513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,2560,0.02221600049071842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,3072,0.09630133046044244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,3072,0.13517599635654026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,2048,0.01789777808719211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,2560,0.08169244395362006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,2560,0.11648178100585938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,1536,0.014904888139830695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,2048,0.06597422228919135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,2048,0.09733244445588853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,1024,0.010689777632554373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,1536,0.05233866638607449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,1024,0.03679733475049337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,768,0.008764444126023186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,1024,0.058541331026289195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,768,0.030858665704727173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,768,0.05099644594722324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,512,0.007271111011505127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,512,0.041167110204696655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,512,0.02423022190729777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,256,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,256,0.01922044489118788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,256,0.03191377719243368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,128,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,128,0.017514665921529133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,1536,0.07848444249894884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,2048,128,0.027225777506828308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,64,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,2048,32,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,64,0.01682844426896837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,2048,32,0.016862221890025668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,65536,0.3685404459635417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,51200,0.2850320074293348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,65536,1.9247307247585719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,16384,0.09741333458158706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,51200,1.5115057627360027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,16384,0.48821157879299587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,12288,0.06932977835337321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,65536,2.6209244198269315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,16384,0.6750177807278104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,10240,0.07151199711693658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,12288,0.3693644470638699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,51200,2.05622312757704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,10240,0.3104577859242757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,8192,0.057321780257754855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,10240,0.43059823248121476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,7168,0.04999377661281162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,8192,0.25018665525648326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,8192,0.34923556115892196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,6144,0.041814221276177295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,7168,0.2193004422717624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,7168,0.3082337909274631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,5120,0.03607733382119073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,6144,0.18904711140526664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,6144,0.2657431231604682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,12288,0.5118977758619521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,4096,0.029356443219714697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,5120,0.15525155597262913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,3584,0.02612888813018799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,4096,0.12082933055029975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,4096,0.17468089527553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,5120,0.22012621826595732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,3072,0.022633777724372015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,3584,0.10476444164911906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,2560,0.019702222612169053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,3584,0.15421066019270155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,3072,0.13468621836768255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,3072,0.0910595522986518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,2048,0.016297777493794758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,2560,0.07702489031685723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,1536,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,2560,0.11621688471900092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,2048,0.09679555230670506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,2048,0.06367111206054688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,1024,0.009639110830095079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,1536,0.04796888762050205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,1536,0.07776088847054376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,768,0.008250666989220513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,1024,0.0574151078859965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,512,0.006583111153708563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,1024,0.034323556555642024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,768,0.048970666196611196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,256,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,512,0.022703111171722412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,512,0.03963466816478305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,256,0.01795377830664317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,128,0.004954666727119022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,256,0.030941334035661485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,128,0.01591822173860338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,64,0.004632888982693355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1536,128,0.026183111800087824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1536,32,0.004774222357405556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,64,0.01554666625128852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,65536,0.25175023078918457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,32,0.015855110353893705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1536,768,0.02884711159600152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,51200,0.20730577574835884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,51200,1.4063955942789714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,65536,1.7922604878743489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,16384,0.06461511055628459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,16384,0.4575804604424371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,65536,2.495346705118815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,12288,0.050981332858403526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,16384,0.6430071194966634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,12288,0.3466586536831326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,10240,0.04157688882615831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,51200,1.9527689615885417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,12288,0.48851198620266384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,8192,0.03312266535229153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,10240,0.29099644554985893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,7168,0.02910311023394267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,10240,0.41104888916015625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,8192,0.23489422268337676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,6144,0.025812443759706285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,8192,0.3328559928470188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,7168,0.20656977759467232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,7168,0.29367465443081325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,5120,0.021372444099850122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,6144,0.17681599987877739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,6144,0.2534693347083198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,4096,0.018379555808173288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,5120,0.14580889542897543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,3584,0.01574133336544037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,4096,0.11353066232469346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,5120,0.20980800522698295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,4096,0.16720799605051676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,3072,0.014750222365061441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,3584,0.09775022003385757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,3584,0.14702844619750977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,2560,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,3072,0.08468711376190186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,2048,0.010913777682516309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,3072,0.12883910867902967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,2560,0.07210666603512235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,2560,0.11027555333243476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,1536,0.00870666652917862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,2048,0.05879111422432793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,1024,0.007391110890441471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,2048,0.09311378002166748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,1536,0.04452000061670939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,1536,0.07421422004699707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,768,0.006663110935025745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,1024,0.03154755632082621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,1024,0.05610489183002048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,768,0.027056000298923914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,768,0.046253333489100136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,256,0.004886222382386525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,512,0.0376835564772288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,512,0.02161955502298143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,256,0.016530666086408827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,128,0.004543999830881755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,256,0.02950577934583028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,64,0.0038515557017591265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,128,0.014480888843536377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,1024,128,0.025117332736651104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,32,0.0041831110914548235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,64,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,65536,0.20980444219377306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,1024,32,0.014436443646748861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,51200,0.1591120031144884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,1024,512,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,65536,1.743576897515191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,51200,1.3683368894788954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,16384,0.052541332112418286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,16384,0.44578308529324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,12288,0.04097955425580343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,51200,1.9543289608425563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,16384,0.6421493424309624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,65536,2.493162579006619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,10240,0.034303112162484065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,12288,0.3382213380601671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,8192,0.02738933265209198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,10240,0.2835466596815321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,10240,0.410354667239719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,7168,0.023769777682092454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,8192,0.22919376691182455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,8192,0.33239110310872394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,7168,0.20088533560434976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,6144,0.1721644401550293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,7168,0.2941475444369846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,5120,0.018595554762416415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,6144,0.2532453272077772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,12288,0.48813597361246747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,4096,0.01591022147072686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,5120,0.14200266202290854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,5120,0.20847911304897734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,4096,0.11043644613689846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,6144,0.0217866665787167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,3584,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,4096,0.16700355211893717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,3072,0.012429333395428128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,3584,0.09560355875227187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,3584,0.14740710788302952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,3072,0.08310133218765259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,2560,0.01110577748881446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,3072,0.12818843788570827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,2048,0.00980888886584176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,2560,0.0700311130947537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,2560,0.10990133550431992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,2048,0.056743999322255455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,1536,0.008405333591832055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,2048,0.09190489186180963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,1024,0.006686222222116258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,1536,0.04259022076924642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,1536,0.07442310783598158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,768,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,1024,0.05429600013626946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,768,0.04640266630384657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,768,0.026438222991095647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,512,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,512,0.037364443143208824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,512,0.020961778031455144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,256,0.015966221690177917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,1024,0.030900445249345567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,256,0.02959200077586704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,128,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,768,128,0.025050666597154405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,64,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,128,0.014206222361988492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,32,0.004173333446184794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,64,0.013830222189426422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,768,256,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,768,32,0.01349244432316886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,65536,0.16261866357591417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,51200,0.13179288970099554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,51200,1.307023154364692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,65536,1.6670221752590602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,16384,0.04144355654716492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,16384,0.4266764322916667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,12288,0.03141955534617106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,16384,0.6087386873033311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,65536,2.3645945654975042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,10240,0.02492888934082455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,12288,0.3234231207105848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,12288,0.4619813495212131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,8192,0.018237334158685472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,51200,1.8519556257459853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,10240,0.2721751001146105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,10240,0.38931822776794434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,7168,0.01646133263905843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,8192,0.21919022666083443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,6144,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,7168,0.19282844331529406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,8192,0.31559377246432835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,7168,0.278472900390625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,5120,0.012960889273219638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,6144,0.16521510812971327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,4096,0.010400888820489248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,6144,0.24016088909573027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,5120,0.13608266247643366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,4096,0.10589333375295003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,5120,0.1987511052025689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,3584,0.009887999958462184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,4096,0.15757866700490317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,3072,0.008985777695973715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,3584,0.0913155542479621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,3584,0.13946221934424505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,2560,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,3072,0.1210906638039483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,2048,0.0069466663731469055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,2560,0.10381778081258138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,2560,0.06666488779915704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,1536,0.006184000107977126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,2048,0.05436177717314827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,2048,0.08711644675996567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,1024,0.005223111146026188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,1536,0.040927112102508545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,1536,0.0684808890024821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,1024,0.029467556211683486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,1024,0.05139288968510098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,768,0.004928888960017098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,512,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,768,0.025103110406133864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,768,0.04408355553944906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,256,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,3072,0.07869955566194323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,512,0.01996977792845832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,512,0.03556177682346768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,128,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,256,0.0277786652247111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,256,0.015190222197108798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,128,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,64,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,512,32,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,512,128,0.02372177773051792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,64,0.012827555338541666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,512,32,0.012737777498033313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,65536,0.09747911161846584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,51200,0.07808533642027113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,51200,1.2732462353176541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,65536,1.6269262101915147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,16384,0.031911111540264554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,16384,0.41744444105360246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,65536,2.3494675954182944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,12288,0.023397333092159692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,12288,0.3157422277662489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,16384,0.6062435574001735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,10240,0.019161777363883126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,12288,0.4598266813490126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,10240,0.26556889216105145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,8192,0.015672889020707872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,8192,0.2145600053999159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,10240,0.38699645466274685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,7168,0.013781333135233985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,8192,0.31410132514105904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,7168,0.18831911351945665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,6144,0.012264000044928657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,51200,1.843123541937934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,7168,0.27688799964057076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,6144,0.16124355792999268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,5120,0.010880000061459012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,4096,0.008716444174448649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,5120,0.13223200374179417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,6144,0.23903200361463758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,5120,0.19752711719936797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,3584,0.008147555920812819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,4096,0.10299111074871486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,3072,0.0070257774657673305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,3584,0.08885688914193048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,3072,0.07659022013346355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,2560,0.006598222172922558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,3072,0.12036977873908149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,3584,0.13799556096394858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,2048,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,2560,0.0647617777188619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,2560,0.1031457781791687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,1536,0.00545155546731419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,2048,0.05149066779348585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,2048,0.08597066667344834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,1024,0.004549333204825719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,1536,0.03897155655754937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,1536,0.06993777884377374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,1024,0.028524445162879095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,768,0.004229333251714706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,1024,0.050740443997912936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,4096,0.1570355494817098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,512,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,768,0.024373332659403484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,512,0.018976888722843595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,256,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,768,0.04390311241149902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,512,0.035248888863457575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,128,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,256,0.027506666051016912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,256,0.014448889427714877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,64,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,128,0.013022222452693515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,256,32,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,256,128,0.023455111516846552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,64,0.012170666621790992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,65536,0.0860826638009813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,256,32,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,51200,0.06454488966200086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,51200,1.2442666159735787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,16384,0.026712889472643535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,65536,1.5884443918863933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,16384,0.406968010796441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,51200,1.840743170844184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,12288,0.01793599956565433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,65536,2.3491795857747397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,16384,0.6062480078803169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,12288,0.3083164427015516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,10240,0.014767999450365702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,8192,0.012088889049159156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,12288,0.46032354566786027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,10240,0.25964710447523326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,8192,0.2091368834177653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,10240,0.3872213363647461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,7168,0.011155555645624796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,8192,0.31328622500101727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,6144,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,7168,0.18393777476416695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,5120,0.00867377801073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,7168,0.27589334381951225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,6144,0.15783645047081843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,4096,0.0076044441925154785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,6144,0.2385608885023329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,5120,0.12903111510806614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,5120,0.19656266106499565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,3584,0.006986666884687211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,4096,0.10024711158540513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,3072,0.006310222049554189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,4096,0.15680800543891057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,3584,0.08658755487865871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,3584,0.1369582282172309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,2560,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,3072,0.07432355483373006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,3072,0.12006666925218369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,2048,0.005689777847793367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,2560,0.1025422215461731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,2048,0.050168000989490084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,2048,0.08591022094090779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,1536,0.004927999857399198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,1536,0.037372443411085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,1024,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,1536,0.06812177764044867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,1024,0.027484445108307734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,768,0.003931555483076307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,1024,0.05111555589569939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,2560,0.06272800101174249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,768,0.02347911066479153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,512,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,768,0.043657776382234365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,256,0.003277333246337043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,512,0.018584888842370775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,512,0.034655110703574285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,128,0.0029475556479560006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,256,0.026848889059490625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,256,0.014242667290899487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,128,0.012231999801264869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,64,0.0029804443733559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,128,32,0.0032559999575217566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,64,0.011789333489206104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,65536,0.07698400153054132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,128,32,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,51200,0.065102219581604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,51200,1.2446969350179036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,16384,0.025490666429201763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,16384,0.40737777286105686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,12288,0.01369866645998425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,10240,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,12288,0.30858132574293345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,10240,0.2589520083533393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,65536,1.5881350835164387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,8192,0.010673777924643623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,7168,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,6144,0.009709333380063375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,8192,0.20938666661580405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,5120,0.008344888687133789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,7168,0.18391911188761392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,6144,0.1573679976993137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,4096,0.007611555357774098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,3584,0.007048000064161088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,5120,0.12916799386342367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,3072,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,4096,0.09979644748899673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,2560,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,3584,0.08668266402350532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,2048,0.0053226666318045715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,2560,0.0625680022769504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,1536,0.004558222161398994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,2048,0.050350222322675914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,1024,0.0041928887367248535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,1536,0.037793778710895114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,768,0.0038408889538711975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,1024,0.027855111493004694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,512,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,512,0.018951111369662814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,256,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,3072,0.07414488659964667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,128,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,256,0.014135110709402295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,128,0.012057777908113269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,64,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,32,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,768,0.023385778069496155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,64,32,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,65536,0.07895733250512017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,51200,0.05930844280454847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,64,64,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,16384,0.023451555106374953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,51200,1.2430737813313801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,12288,0.01073333372672399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,16384,0.406912883122762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,65536,1.586716439988878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,10240,0.00980888886584176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,8192,0.00907466643386417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,12288,0.30813688702053493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,7168,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,10240,0.25891733169555664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,6144,0.008653333617581261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,8192,0.20946044392055937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,7168,0.1834391090604994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,6144,0.15782666206359863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,5120,0.010771555205186209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,4096,0.008981333010726506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,3584,0.008664888640244802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,5120,0.12915555636088052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,3072,0.0075937774446275495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,3584,0.08624444405237834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,4096,0.09981866677602132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,2560,0.00693866651919153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,2048,0.0052266667286554975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,3072,0.07432355483373006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,2560,0.06343022320005628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,2048,0.050048000282711454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,1536,0.004690666579537922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,1024,0.003946666502290302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,1536,0.03726666503482395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,768,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,1024,0.027836445305082533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,768,0.023669333921538457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,512,0.01867999964290195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,256,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,256,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,128,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,64,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,512,0.003495111233658261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,64,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3584,32,32,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3584,32,32,0.011707555916574268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,16384,3.471211539374458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,51200,10.521848890516493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,16384,4.860705057779948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,12288,2.6827394697401257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,12288,3.3545892503526478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,10240,2.109295103285048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,51200,16.68947516547309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,8192,1.6902817620171442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,10240,2.743736902872721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3584,128,128,0.022300443715519373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,7168,1.473854170905219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,8192,2.1052773793538413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,6144,1.2852346632215712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,7168,1.8423663245307074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,5120,1.0634382035997179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,6144,1.5652390586005316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,4096,0.8570835325453016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,5120,1.3004248936971028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,3584,0.7872186766730415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,4096,1.0330702463785808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,3584,0.9021413591172961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,3072,0.6770266426934136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,2560,0.5383297602335612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,2048,0.44542577531602645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,2560,0.6607315805223253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,1536,0.33209511968824595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,3072,0.7797564400566949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,2048,0.5386853218078613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,1024,0.23533333672417533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,1536,0.424377785788642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,768,0.17915022373199463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,512,0.15214044517940944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,1024,0.3212044503953722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,256,0.11079644494586521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,768,0.2543831136491564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,512,0.21946844789716932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,128,0.0844622254371643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,64,0.08366933133867051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,65536,32,0.08581066793865627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,128,0.1821671062045627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,64,0.18552445040808785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,256,0.19619377454121908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,65536,32,0.1892497804429796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,51200,8.190706888834635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,65536,10.622662862141928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,16384,2.8102684020996094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,12288,2.0355076260036893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,51200,13.27952660454644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,16384,3.8701883951822915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,10240,1.654873741997613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,12288,2.699142244127062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,8192,1.3210915459526908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,10240,2.1859963734944663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,65536,17.232779608832466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,7168,1.175305790371365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,8192,1.6958693398369684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,6144,1.0495351155598958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,7168,1.4827697541978624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,5120,0.8571919865078397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,4096,0.6716693242390951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,6144,1.2515271504720051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,5120,1.0408702426486545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,3584,0.6000017589992946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,3072,0.5215377807617188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,3584,0.7232577535841199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,4096,0.8280808660719129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,2560,0.45566309822930234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,3072,0.625440862443712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,2048,0.34679733382330996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,2560,0.5300141970316569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,1536,0.2850488821665446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,1024,0.184517330593533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,2048,0.432370662689209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,768,0.14321155018276638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,1024,0.25716177622477215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,1536,0.34276533126831055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,768,0.2040319972568088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,256,0.0867875549528334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,512,0.17329955101013184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,128,0.06508177518844604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,64,0.05553244551022848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,128,0.14531021647983128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,64,0.14849155479007298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,32,0.058541331026289195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,51200,512,0.1077573365635342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,32,0.1509528954823812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,51200,256,0.153766221470303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,51200,2.461896048651801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,65536,3.214932547675239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,16384,0.8582320213317871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,12288,0.6193688710530599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,16384,1.4494462543063695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,10240,0.5185102356804742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,12288,1.0358346303304036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,65536,6.316412607828776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,8192,0.41747376653883195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,10240,0.8460870848761665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,7168,0.38074400689866805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,8192,0.6676328976949056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,6144,0.3340551058451335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,7168,0.5831662284003364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,5120,0.29710756407843697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,51200,5.036739349365234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,4096,0.2278684510125054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,5120,0.41257421175638836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,3584,0.19875200589497885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,4096,0.3250959979163276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,3072,0.17062577936384413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,3584,0.28500088055928546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,2560,0.13863378100925022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,3072,0.24676444795396593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,2048,0.12529688411288792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,2560,0.21111822128295898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,1536,0.09476711352666219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,2048,0.17336266570621064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,1024,0.06772266493903266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,768,0.05278933379385206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,1536,0.1373280021879408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,6144,0.49943913353814023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,512,0.03932355509863959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,256,0.03014311194419861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,768,0.08365599976645575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,128,0.024711999628278945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,256,0.060328000121646456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,512,0.07038044267230563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,64,0.020977778567208186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,128,0.0561751127243042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,16384,32,0.021228444245126512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,32,0.05767911010318332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,64,0.057272891203562416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,16384,1024,0.10330488946702744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,51200,2.0386106703016496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,65536,2.6351085238986545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,16384,0.6559368769327799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,12288,0.5044630898369683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,16384,1.1659990946451824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,10240,0.42879732449849445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,12288,0.8374142116970487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,51200,4.054676479763455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,8192,0.34940621587965226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,10240,0.6970106760660807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,7168,0.2958995501200358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,8192,0.5467848777770996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,7168,0.478931532965766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,6144,0.2662373383839925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,5120,0.21873511208428276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,6144,0.4090035491519504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,4096,0.1691582202911377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,5120,0.3500657876332601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,3584,0.15485333071814641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,4096,0.2662622133890788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,3072,0.1263200044631958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,2560,0.11016711261537339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,65536,5.223993089463976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,3072,0.20249333646562365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,3584,0.23445243305630156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,2048,0.09559377696779038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,1536,0.06672533353169759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,2560,0.17326400015089247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,1024,0.0532551109790802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,2048,0.1427075597974989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,768,0.03826844361093309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,1024,0.08523022466235691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,768,0.06945333215925428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,256,0.022828444838523865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,512,0.058005332946777344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,128,0.01904000010755327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,256,0.048912889427608915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,64,0.01681155628628201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,128,0.04483200112978617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,32,0.017177777157889474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,64,0.04532444477081299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,32,0.046259555551740855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,12288,1536,0.11319377687242295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,12288,512,0.02923111120859782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,51200,1.614307509528266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,65536,2.0743804507785373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,16384,0.5511573155721029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,12288,0.39730132950676816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,16384,1.017173343234592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,10240,0.33858222431606716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,12288,0.7394853168063693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,51200,3.530760871039497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,10240,0.6126666598849827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,8192,0.2857866552140978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,65536,4.483111911349826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,7168,0.2378293408287896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,8192,0.48643287022908527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,6144,0.20615200201670328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,7168,0.42693156666225857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,5120,0.1789217789967855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,6144,0.3647475507524278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,4096,0.1422213315963745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,3584,0.12607378429836696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,4096,0.2368533346388075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,3072,0.11052622397740682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,3584,0.20795733398861357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,2560,0.08771910932328965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,2560,0.15494933393266466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,5120,0.3006213241153293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,2048,0.07159111234876844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,1536,0.05594311157862345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,2048,0.12715644306606716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,1024,0.039990223116344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,1536,0.10169155730141534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,3072,0.18054489294687906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,768,0.03205688794453939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,512,0.023424888650576275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,1024,0.07632622453901503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,256,0.018901333212852478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,768,0.06189155578613281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,128,0.01632355484697554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,256,0.04280266828007168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,512,0.051551110214657254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,64,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,128,0.03803733322355483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,10240,32,0.014158222410413953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,32,0.03679200013478597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,10240,64,0.03751733236842685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,51200,1.2684995863172743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,65536,1.6647680070665147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,16384,0.4102053377363417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,12288,0.3114657666948107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,16384,0.8868480258517795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,10240,0.2796853383382161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,12288,0.6413928667704264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,8192,0.2115759982003106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,10240,0.5337448649936253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,7168,0.1991715563668145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,51200,3.0712782541910806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,6144,0.1841120057635837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,65536,3.955145094129774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,7168,0.37448710865444607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,5120,0.14772710535261366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,6144,0.32040709919399685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,5120,0.2636471059587267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,3584,0.09889777501424153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,4096,0.2065840032365587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,3072,0.08624977535671657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,3584,0.18258844481574166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,2560,0.07239377498626709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,8192,0.42541156874762637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,2048,0.06368266873889498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,2560,0.1359822220272488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,1536,0.04559555649757385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,4096,0.11563200420803493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,2048,0.11242577764723037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,1024,0.03375466664632162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,1536,0.08903377585940891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,768,0.02536711096763611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,512,0.01998488936159346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,768,0.05476266807980008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,1024,0.0669635534286499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,512,0.04324088825119866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,128,0.013564444250530668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,256,0.035094221433003746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,128,0.03186755710177951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,3072,0.1590826643837823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,64,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,32,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,64,0.031530665026770696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,8192,32,0.03151911165979173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,8192,256,0.015840889679061044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,65536,1.530184851752387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,51200,1.2234790590074327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,16384,0.4136959976620144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,12288,0.30509599049886066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,16384,0.8187422222561307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,10240,0.24471378326416016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,12288,0.6129191186692979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,8192,0.19759910636477998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,10240,0.4963582356770833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,51200,2.7982879214816623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,65536,3.6263198852539062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,7168,0.1706871059205797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,8192,0.39523998896280926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,5120,0.12423288822174072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,6144,0.1461662186516656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,7168,0.3483146561516656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,6144,0.2977955606248644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,5120,0.24465510580274794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,4096,0.19209244516160753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,3584,0.0891466670566135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,3072,0.07905777957704332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,3584,0.16838488313886854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,2560,0.06345422400368585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,3072,0.14772266811794704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,2560,0.12620622581905788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,2048,0.05243644449445936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,1536,0.040236443281173706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,4096,0.0966435538397895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,1024,0.03173333406448364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,2048,0.10428800185521443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,768,0.02509244448608822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,1536,0.08348800076378717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,512,0.01947111056910621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,1024,0.06170044342676798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,256,0.014743111199802823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,256,0.03147911032040914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,512,0.03941955500178867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,128,0.0159262220064799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,64,0.014134221606784396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,128,0.029508444997999404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,64,0.029116445117526587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,7168,32,0.014475555883513557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,32,0.028656889994939167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,7168,768,0.049959109889136426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,65536,1.2827360365125868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,51200,0.983159065246582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,16384,0.3453271124098036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,16384,0.7515689002143012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,12288,0.2590266598595513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,10240,0.20494400130377877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,65536,3.3397901323106556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,12288,0.5504959954155816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,8192,0.17844977643754747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,7168,0.15416444672478571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,10240,0.4568159845140245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,51200,2.5631752014160156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,6144,0.12735999955071345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,8192,0.36565423011779785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,5120,0.10360888640085857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,4096,0.08331111404630873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,7168,0.334278212653266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,6144,0.2757493389977349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,5120,0.2263688776228163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,3072,0.0649075574345059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,4096,0.1773244407441881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,2560,0.05623822079764473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,3584,0.15622844960954455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,3072,0.13497511545817056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,2048,0.04581155710750156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,2560,0.11681511667039658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,1536,0.03466755482885573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,1024,0.026477333572175767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,2048,0.09650933080249363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,768,0.02091111077202691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,1536,0.0765626695421007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,512,0.015232889188660515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,768,0.045479999648200146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,256,0.012303111453851065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,1024,0.05651288562350803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,512,0.0361422234111362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,3584,0.08102311028374566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,128,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,64,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,6144,32,0.011352889239788055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,256,0.028840889533360798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,128,0.026743110683229234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,64,0.026750221848487854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,6144,32,0.026131555438041687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,51200,0.8326764636569552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,16384,0.2929351064893934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,16384,0.6935244666205512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,65536,1.028082635667589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,12288,0.22217067082722983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,51200,2.340575959947374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,10240,0.18167466587490508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,12288,0.5044639905293783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,8192,0.14840266439649794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,65536,2.934332529703776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,10240,0.41907821761237246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,7168,0.1337217754787869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,6144,0.10553599728478326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,8192,0.33597776624891496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,5120,0.09413955609003703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,7168,0.2954382101694743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,4096,0.0736231141620212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,6144,0.2538088957468669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,3584,0.06545866860283746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,3072,0.05724000268512302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,4096,0.16222577624850804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,2560,0.04853777752982246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,3072,0.12358578046162923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,3584,0.14325688944922552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,2048,0.03832888934347365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,2560,0.10558578040864731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,1536,0.029560890462663438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,1024,0.02201244400607215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,1536,0.07015377945370145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,2048,0.08904800150129531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,768,0.017794667018784415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,1024,0.05108799868159824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,512,0.013074666261672974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,768,0.042318221595552236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,256,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,5120,0.20792088243696424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,128,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,256,0.02647022240691715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,64,0.008314666648705801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,128,0.024298666252030268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,5120,32,0.008316444853941599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,32,0.023741333021057978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,64,0.024476443727811176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,5120,512,0.032460444503360324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,65536,0.8273404439290365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,51200,0.640499538845486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,16384,0.2330711152818468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,16384,0.6228444311353896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,12288,0.17812889152102998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,65536,2.6390772925482855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,12288,0.4621111022101508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,10240,0.14728444152408177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,51200,2.082381354437934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,8192,0.1216542190975613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,7168,0.10767289002736409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,10240,0.38272444407145184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,8192,0.30654578738742405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,6144,0.08850400315390693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,5120,0.07854133182101779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,4096,0.061307556099361844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,7168,0.2702088885837131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,6144,0.2314008871714274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,3584,0.05109688970777723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,3072,0.0464053319560157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,4096,0.14805332819620767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,2560,0.03826577795876397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,3584,0.13053422504001194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,2048,0.031672000885009766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,2560,0.09647733635372585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,3072,0.1134106715520223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,1536,0.024470221665170457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,2048,0.07810578081342909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,1024,0.02037244372897678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,768,0.01608533329433865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,1024,0.045978665351867676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,1536,0.06388888756434123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,512,0.011526222030321756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,768,0.037490665912628174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,256,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,128,0.007642666498819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,256,0.023756444454193115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,64,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,512,0.0295688874191708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,128,0.02183733383814494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,4096,32,0.007333333293596904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,5120,0.18985511196984184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,64,0.02161866592036353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,4096,32,0.021337777376174927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,51200,0.58824708726671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,65536,0.7668079800075955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,16384,0.2113093270195855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,12288,0.15359289116329616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,16384,0.586176872253418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,12288,0.44997511969672305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,10240,0.14370044072469076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,8192,0.11965955628289117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,51200,1.9555394914415147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,10240,0.3642711109585232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,7168,0.09545422262615627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,6144,0.08657422330644395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,8192,0.2927475505405002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,65536,2.466064876980252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,7168,0.25744888517591685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,5120,0.06855110989676581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,4096,0.057012445396847196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,3584,0.05136977632840475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,6144,0.22032177448272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,5120,0.18078577518463135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,3072,0.044778668218188815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,4096,0.14181689421335855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,2560,0.03815644317203098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,2048,0.02853511108292474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,3584,0.12401599354214138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,3072,0.10733155409495036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,1536,0.023437332775857713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,1024,0.015863999724388123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,2048,0.07382577657699585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,2560,0.09174488650427924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,1536,0.059579557842678495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,768,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,1024,0.04312710960706075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,512,0.010788444843557147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,256,0.008300444318188561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,128,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,512,0.027857777145173814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,256,0.022671999202834234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,768,0.0364586677816179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,64,0.0069208890199661255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,128,0.020649777518378366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3584,32,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,64,0.020311999652120803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3584,32,0.020600888464185927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,65536,0.5736391279432509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,51200,0.483284420437283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,16384,0.178802662425571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,16384,0.5896053314208984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,12288,0.13109421730041504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,65536,2.2506391737196183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,51200,1.8031804826524522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,10240,0.10283022456698948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,8192,0.08230133189095391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,10240,0.3446017901102702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,7168,0.07508711020151775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,8192,0.2777644528283013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,6144,0.06682311164008246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,7168,0.24402311113145617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,12288,0.411448876063029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,5120,0.05444444550408257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,4096,0.04440266556209988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,6144,0.20792799525790742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,3584,0.03867644402715895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,5120,0.17047288682725695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,3072,0.03372800019052293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,4096,0.13374222649468318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,2560,0.029468443658616807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,3584,0.11807999345991348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,3072,0.1017448902130127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,2048,0.024925332930352952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,2560,0.08599288596047296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,1536,0.02141155633661482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,1024,0.01605422298113505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,2048,0.0694426695505778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,768,0.012983111043771109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,1536,0.05353866683112251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,512,0.010429333481523724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,1024,0.041101333167817854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,768,0.033371554480658636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,256,0.008648888932334052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,512,0.025808889004919264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,64,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,128,0.007252444823582967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,3072,32,0.006630222416586346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,256,0.020981333321995203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,64,0.018904000520706177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,128,0.019539554913838703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,3072,32,0.019301333361201815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,65536,0.5278462304009331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,51200,0.4135084417131212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,16384,0.14137066735161677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,12288,0.11109511057535808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,16384,0.5148728688557943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,10240,0.0876622200012207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,12288,0.38955733511183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,8192,0.07368799712922838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,10240,0.32546665933397084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,7168,0.06239022148980034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,8192,0.2627600034077962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,51200,1.6378862592909071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,6144,0.05451110998789469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,5120,0.04574133290184868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,7168,0.23078844282362196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,65536,2.127040015326606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,4096,0.03731200098991394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,6144,0.19681510660383436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,3584,0.03271199928389655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,5120,0.1616551081339518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,3072,0.02881866693496704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,2560,0.025080889463424683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,4096,0.12707111570570204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,2048,0.021879111727078755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,3072,0.09639733367496067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,1536,0.017743999759356182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,3584,0.11206400394439697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,2560,0.08109688758850098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,1024,0.012441777520709567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,2048,0.06599111027187772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,768,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,1536,0.05002311203214857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,512,0.008422222402360704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,1024,0.03805244300100539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,256,0.006963555183675554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,768,0.03169599837727017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,512,0.02478666603565216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,256,0.019928000039524503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,64,0.005568888866239124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,32,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,128,0.01799199978510539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,64,0.01793333391348521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2560,32,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,65536,0.40994755427042645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,51200,0.34908000628153485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,16384,0.11813510788811578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2560,128,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,12288,0.08403199911117554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,16384,0.5089155303107368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,10240,0.07700266440709432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,12288,0.3683004379272461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,8192,0.062011553181542285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,10240,0.30941422780354816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,7168,0.051669332716200084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,51200,1.5330133438110352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,65536,1.9549929300944011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,6144,0.046569777859581836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,8192,0.24947910838656953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,5120,0.03825866513782077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,7168,0.21960800223880342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,4096,0.032469332218170166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,6144,0.18674311372968885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,3584,0.028373334142896865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,4096,0.11988533867730035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,5120,0.1531368891398112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,3072,0.025256888733969793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,2560,0.021173333128293354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,3584,0.10509600241978963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,2048,0.01829333272245195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,3072,0.0915599995189243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,2560,0.07679466406504314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,2048,0.06407555606630114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,1536,0.013818666338920593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,1024,0.01034311122364468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,768,0.00869066682126787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,1536,0.049345778094397656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,512,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,1024,0.035087111923429705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,768,0.030717333157857258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,256,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,128,0.005321777943107817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,512,0.02346755564212799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,64,0.0051893335249688895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,256,0.018627555833922494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,2048,32,0.004959110998445087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,128,0.01718755563100179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,64,0.01650311052799225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,2048,32,0.016545777519543965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,65536,0.3070800039503309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,51200,0.24287199974060059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,16384,0.08930933475494385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,16384,0.4589342011345758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,12288,0.0711395541826884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,12288,0.34652267562018496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,10240,0.050809777445263334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,51200,1.422009785970052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,65536,1.8072861565483942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,8192,0.041413333680894636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,7168,0.03680355681313409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,10240,0.2913911077711317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,8192,0.23475376764933267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,6144,0.03232977787653605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,5120,0.02713866697417365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,7168,0.20622489187452528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,4096,0.02313866714636485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,6144,0.17627022001478407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,5120,0.14440355035993788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,3584,0.021684444612926904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,3072,0.018951111369662814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,4096,0.11331466833750407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,3072,0.08550133307774861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,2560,0.07299555672539605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,2048,0.014753777119848462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,1536,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,2048,0.05899822049670749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,1024,0.009346666435400644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,1536,0.0452186663945516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,768,0.008005333443482717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,1024,0.03257955445183648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,512,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,768,0.02785244584083557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,256,0.005722666780153911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,3584,0.09882933563656277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,512,0.02200800014866723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,128,0.004906666775544484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,2560,0.016168889072206285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,64,0.004567999806669024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,256,0.017254221770498488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1536,32,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,128,0.01553600033124288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,64,0.015447111593352424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,65536,0.22398222817314994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,51200,0.17942133214738634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,16384,0.06015021933449639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,16384,0.429512898127238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,12288,0.04353244437111748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1536,32,0.015107555521859063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,10240,0.03698755635155572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,51200,1.3164151509602864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,8192,0.0328382220533159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,12288,0.3256426652272542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,65536,1.6782666312323675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,7168,0.02700177828470866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,10240,0.27292709880405003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,6144,0.02330577704641554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,5120,0.019464888506465487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,8192,0.22030933698018393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,4096,0.016520000166363187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,7168,0.19310488965776232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,3584,0.015055111712879606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,6144,0.1648728847503662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,3072,0.013658666776286231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,4096,0.10524621937010024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,5120,0.1344719992743598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,2560,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,3584,0.09234133031633164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,2048,0.00999111102686988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,3072,0.07979733414120145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,1536,0.008336889247099558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,2560,0.06814844740761651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,2048,0.054766221178902515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,1536,0.041703111595577665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,768,0.005954666684071223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,1024,0.030211554633246526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,512,0.0052915554907586836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,768,0.026018665896521673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,256,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,512,0.020619556307792664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,128,0.004192000048028098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,256,0.016226665841208566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,64,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,128,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,32,0.003964444415436851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,64,0.01422044469250573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,1024,32,0.014188443620999655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,1024,1024,0.007277333074145847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,65536,0.19110222657521567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,51200,0.1517911089791192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,16384,0.05248533354865181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,16384,0.4187280072106256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,12288,0.04009333252906799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,51200,1.2788195080227323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,10240,0.03270755542649163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,12288,0.31755023532443577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,8192,0.02665600015057458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,65536,1.6326124403211806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,7168,0.02365422248840332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,10240,0.2661626603868273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,6144,0.021183111601405676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,8192,0.21506933371225992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,7168,0.1881564458211263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,5120,0.01797066628932953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,4096,0.015544889701737298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,6144,0.16038044293721518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,3584,0.014059555199411182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,5120,0.13123644722832575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,3072,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,4096,0.10286221901575725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,3584,0.0895342230796814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,2560,0.011152000062995486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,2048,0.009507555928495195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,3072,0.07747911082373725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,1536,0.008380444513426887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,2560,0.0660382244322035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,1024,0.006664000037643645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,2048,0.05340533455212911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,1536,0.04045333464940389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,768,0.006076444354322221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,1024,0.029570665624406602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,512,0.004936888813972473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,256,0.004734222259786394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,768,0.025205334027608235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,512,0.019928000039524503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,64,0.004078222231732475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,128,0.013604444762070974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,256,0.015655999382336933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,32,0.0041982221106688184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,64,0.013420444395807056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,65536,0.14175111717647976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,768,32,0.01348444405529234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,51200,0.11597155200110541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,16384,0.04121422105365329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,16384,0.39959200223286945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,768,128,0.00425155543618732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,12288,0.02934933371014065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,10240,0.0232933329211341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,12288,0.3031448788113064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,10240,0.2549804316626655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,51200,1.2216471566094291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,7168,0.01683466633160909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,65536,1.5573422114054363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,6144,0.01481599940194024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,7168,0.1806186702516344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,6144,0.1536346673965454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,5120,0.012888888518015543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,4096,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,5120,0.12618754969702825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,3584,0.009880888793203566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,4096,0.09820977846781413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,3072,0.009039999710188972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,8192,0.019967110620604623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,2560,0.008273777862389883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,3584,0.08505244387520684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,2048,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,3072,0.0743653310669793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,8192,0.20549421840243867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,1536,0.006675555474228329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,2048,0.050055109792285495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,1024,0.005672889037264719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,768,0.004982222285535601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,1536,0.03731733229425218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,512,0.004607111215591431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,1024,0.02808800008561876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,256,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,768,0.02494666642612881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,512,0.018930666976504855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,128,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,256,0.014878221684032016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,64,0.0034924443397257063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,128,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,512,32,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,64,0.012853333519564735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,65536,0.09777333339055379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,32,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,512,2560,0.06250488758087158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,51200,0.07680177688598633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,16384,0.03229866756333245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,16384,0.38956711027357316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,12288,0.021713778376579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,12288,0.2951351006825765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,10240,0.01793333391348521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,51200,1.1871129141913521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,8192,0.014827556080288358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,65536,1.514649815029568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,7168,0.013016888664828407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,10240,0.24785155720180937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,6144,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,8192,0.20037510659959581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,5120,0.010664000279373594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,7168,0.17608355151282418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,6144,0.14968178007337782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,4096,0.008964444200197855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,3584,0.008016000191370646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,5120,0.12222754955291748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,3072,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,4096,0.09468977981143528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,3584,0.08274400234222412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,3072,0.07151822249094646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,2048,0.006248888870080312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,2560,0.06061066521538628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,1536,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,2048,0.048304888937208385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,1024,0.004882666799757215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,1536,0.03630400035116408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,768,0.004187555362780889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,1024,0.027119111683633592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,512,0.003994666867785984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,768,0.02368800010946062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,256,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,512,0.018558222386572096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,128,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,256,0.014478221535682678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,2560,0.0069413334131240845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,64,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,256,32,0.00350222239891688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,128,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,65536,0.0812551114294264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,64,0.012270222107569376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,256,32,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,51200,0.061009777916802295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,16384,0.025158221522967022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,16384,0.37948089175754124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,12288,0.015994666351212394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,10240,0.014212444424629211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,10240,0.24200267261928984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,51200,1.1579280429416232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,8192,0.0120328888297081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,7168,0.010999110837777456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,65536,1.4795697530110676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,12288,0.28870312372843426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,6144,0.00998044427898195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,8192,0.1953199969397651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,5120,0.00869333330127928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,7168,0.1714408927493625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,4096,0.007678222325113084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,3584,0.007615110940403408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,5120,0.1191822157965766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,6144,0.14611732959747314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,4096,0.09205155240164863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,3072,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,2560,0.006261333409282897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,3584,0.08051555686526828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,2048,0.005572444448868434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,1536,0.004920000003443824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,2560,0.05817421939637926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,1024,0.004193777839342753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,2048,0.045943111181259155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,768,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,1536,0.035240888595581055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,512,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,3072,0.07013510995441012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,1024,0.026542221506436665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,256,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,768,0.022787556052207947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,512,0.017892445127169292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,128,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,64,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,256,0.013805333111021252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,128,32,0.003177777760558658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,128,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,65536,0.07312089204788208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,32,0.011713777979214987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,51200,0.05828266673617893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,16384,0.022229333718617756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,128,64,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,16384,0.3792568842569987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,12288,0.013357333011097379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,12288,0.2875351111094157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,10240,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,51200,1.1575181749131944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,8192,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,65536,1.479381349351671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,7168,0.010070222119490305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,10240,0.2415928840637207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,6144,0.009698666632175446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,7168,0.1710186666912503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,5120,0.008732444710201686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,6144,0.14559466309017605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,4096,0.0075839997993575195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,5120,0.11873777707417806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,4096,0.09199200073877971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,3584,0.007308444215191736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,3072,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,3584,0.0804542236857944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,2560,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,3072,0.06903644402821858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,2048,0.005230222311284807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,8192,0.1951671176486545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,2048,0.04547022117508782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,1536,0.0350542234049903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,1024,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,1024,0.026328888204362657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,768,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,768,0.02271911170747545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,512,0.003527111063400904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,512,0.017906667457686532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,2560,0.058196445306142174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,256,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,1536,0.00453955555955569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,128,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,64,0.003148444410827425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,256,0.013761777844693927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,64,32,0.003157333367400699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,128,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,64,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,65536,0.07197778092490302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,64,32,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,51200,0.05854488743676079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,16384,0.021566222111384075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,12288,0.019096889429622226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,16384,0.37929688559638125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,10240,0.016159110599093966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,10240,0.24166843626234266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,12288,0.2877911196814643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,8192,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,7168,0.009310222334331935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,51200,1.1576329337226021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,6144,0.012730666332774691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,65536,1.4790532853868272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,5120,0.011968000067604912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,8192,0.1951448917388916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,7168,0.171287112765842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,6144,0.14568621582455105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,3584,0.008991999758614434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,4096,0.009497777455382878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,3072,0.007978666987684038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,5120,0.11877689096662734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,4096,0.09177866909239028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,2560,0.007600888609886169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,2048,0.005261333452330695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,3584,0.0806657804383172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,1536,0.004548444516128964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,2560,0.05812178055445353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,1024,0.0038284444146686127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,2048,0.04549955659442478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,1536,0.03495288888613383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,768,0.0035546666218174826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,512,0.0032595555401510666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,1024,0.02643644478585985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,256,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,768,0.022806222240130108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,128,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,512,0.01793511046303643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,64,0.0028159999185138275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,256,0.013739555246300168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3329,32,32,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,128,0.011715555356608497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,64,0.01149866647190518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,32,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3329,32,3072,0.07001244359546237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,16384,3.2690747578938804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,16384,4.802002800835504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,51200,9.93857913547092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,12288,2.602713690863715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,16384,3.5507678985595703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,12288,3.293285369873047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,51200,16.39988030327691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,51200,11.400596618652344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,10240,2.2644498613145614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,12288,2.6299866570366754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,10240,2.6904320187038846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,8192,1.6477937698364258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,10240,2.200169881184896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,8192,2.0680035485161676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,7168,1.436095979478624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,7168,1.8084790971544054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,8192,1.745663113064236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,7168,1.5646355946858723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,6144,1.3334613376193578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,6144,1.5324462254842122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,5120,1.0213226742214627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,6144,1.329644415113661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,5120,1.2783306969536674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,4096,0.8557208908928765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,5120,1.1169804467095268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,4096,1.0064853032430012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,3584,0.7275582419501411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,3584,0.8857350879245334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,4096,0.9089590708414713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,3072,0.6037599775526259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,3584,0.8014355765448676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,3072,0.764692465464274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,2560,0.5348258018493652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,3072,0.7036906878153483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,2048,0.43462223476833767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,2560,0.6515102386474609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,2560,0.5997449027167426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,1536,0.3230382336510552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,2048,0.5325093269348145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,2048,0.49999109903971356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,1024,0.21407822767893472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,1536,0.4194435543484158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,1536,0.3981502321031358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,768,0.1880177789264255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,1024,0.3164506753285726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,1024,0.30370577176411945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,512,0.1322071154912313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,768,0.2509537802802192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,768,0.25124621391296387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,256,0.10096444686253865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,512,0.2020275592803955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,128,0.08355466524759929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,512,0.21381778187221953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,256,0.18995822800530326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,256,0.1586151123046875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,64,0.06990755266613431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,65536,32,0.07190222210354276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,128,0.17941778235965303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,64,0.18326132827334932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,65536,128,0.1250800026787652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,65536,32,0.18642755349477133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,51200,8.332939147949219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,65536,10.788499620225695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,65536,16.398030598958332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,65536,11.892779032389322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,16384,2.776823043823242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,16384,3.816727956136068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,51200,13.069185045030382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,12288,1.998768064710829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,16384,2.8791465759277344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,51200,9.12636990017361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,12288,2.628031200832791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,10240,1.7834479014078777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,10240,2.147874620225694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,12288,2.131522708468967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,8192,1.3890640470716689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,10240,1.7669021818372939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,8192,1.6699822743733723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,7168,1.2263156043158638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,8192,1.4302604463365343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,7168,1.4519733852810328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,6144,0.9912684758504232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,7168,1.26036442650689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,5120,0.8186977704366049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,6144,1.2302844789293077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,6144,1.0858195622762044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,4096,0.6226977772182888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,5120,1.0251519944932725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,5120,0.9111048380533854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,4096,0.8070391019185384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,3584,0.5713929070366753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,4096,0.7415297826131185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,3072,0.5011430846320258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,3584,0.708314683702257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,3584,0.6558595763312446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,2560,0.4115368790096707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,3072,0.6148933304680718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,3072,0.5747795634799534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,2048,0.3214435577392578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,2560,0.5224453078375922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,2560,0.4863724708557129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,1536,0.25793777571784127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,2048,0.4267413351270888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,1024,0.18018578158484566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,1536,0.3362880018022325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,1536,0.32842667897542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,768,0.14561511410607234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,1024,0.25400977664523655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,768,0.20440710915459526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,768,0.2058675554063585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,512,0.11199288898044163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,512,0.17066489325629342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,512,0.16668355464935303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,256,0.08013511366314359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,2048,0.4091626538170709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,256,0.12973422474331325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,256,0.15175999535454646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,128,0.06520266665352716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,128,0.14358754952748617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,128,0.10250400172339545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,51200,1024,0.24845155080159506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,64,0.05671733617782593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,51200,32,0.059239996804131397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,64,0.14578399393293592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,51200,32,0.14825422234005398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,65536,3.41304079691569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,51200,2.615049786037869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,65536,6.364763471815322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,65536,4.975646124945746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,16384,0.8443013297186958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,16384,1.414165284898546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,12288,0.6303902202182347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,16384,1.2646453645494249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,51200,4.90484619140625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,10240,0.5327075322469076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,12288,1.0129244062635634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,12288,0.949120839436849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,8192,0.43871378898620605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,10240,0.8368026415506998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,10240,0.7976924578348795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,51200,3.8905245463053384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,8192,0.6764479743109809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,8192,0.6455137994554307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,7168,0.5752711296081543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,6144,0.3013875484466553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,7168,0.570176018608941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,6144,0.4913902282714844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,5120,0.27560533417595756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,7168,0.37124355634053546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,5120,0.40755200386047363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,4096,0.2149404419793023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,6144,0.4929911295572917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,5120,0.41130754682752824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,3584,0.19224266211191812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,4096,0.31997156143188477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,4096,0.3324008782704671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,3584,0.28147466977437335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,3072,0.16295466158125135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,3584,0.2944400045606825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,2560,0.13506310515933567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,3072,0.24398220909966362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,2048,0.10879377524058025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,3072,0.258041779200236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,2560,0.2206231090757582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,1536,0.08457777897516887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,2560,0.20936532815297446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,2048,0.18471199936336943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,2048,0.1716773377524482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,1024,0.06470577584372626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,1536,0.13558844725290933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,1024,0.10222755538092719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,1536,0.14994222588009304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,768,0.04937866661283705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,512,0.03570666578080919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,768,0.08269244432449341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,768,0.09675111373265584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,512,0.06914933522542317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,256,0.027229333917299908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,512,0.07841777801513672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,128,0.02307022280163235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,256,0.059511111842261426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,256,0.06165866719351875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,64,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,128,0.05571111043294271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,128,0.0499875545501709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,16384,32,0.022016889519161646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,64,0.056199113527933754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,16384,32,0.05698577562967936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,16384,1024,0.11377422014872234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,51200,2.1227679782443576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,65536,2.6062329610188804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,65536,4.983869340684679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,65536,4.308597140842013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,16384,0.6794142193264432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,51200,3.988663991292318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,16384,1.143781344095866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,51200,3.376148435804579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,12288,0.5183182292514378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,16384,1.0958693822224934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,10240,0.44584178924560547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,12288,0.8232195642259387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,12288,0.8253297805786133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,8192,0.3548399872250027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,10240,0.6816248893737793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,10240,0.6932097540961372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,7168,0.29643911785549587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,8192,0.5379689004686143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,8192,0.5598115391201443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,6144,0.2562231222788493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,7168,0.47300264570448136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,7168,0.49343024359809023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,5120,0.21666933430565727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,6144,0.40401776631673175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,6144,0.4399884541829427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,4096,0.1683164437611898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,5120,0.33479110399882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,5120,0.35564266310797793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,4096,0.26297956042819554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,3584,0.15501599841647679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,4096,0.28718222512139213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,3584,0.23121865590413412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,3584,0.2526400089263916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,3072,0.13394933276706272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,2560,0.10715022352006699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,3072,0.2004408836364746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,3072,0.2226311100853814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,2048,0.08886488940980698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,2560,0.1912897825241089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,2560,0.1718684434890747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,1536,0.0680968893898858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,2048,0.1413431167602539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,2048,0.15993421607547337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,1024,0.05066755414009094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,1536,0.1292248831854926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,1536,0.11236355039808486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,768,0.03896088732613458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,1024,0.08442399899164836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,1024,0.09874844551086426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,512,0.03007377849684821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,768,0.08369511365890503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,768,0.06908355818854438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,256,0.02575911084810893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,512,0.06858844227261014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,512,0.05664800273047554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,128,0.022359111242824133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,256,0.048862222168180675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,256,0.052961776653925575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,64,0.02104533380932278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,128,0.044940445158216685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,12288,128,0.04454311066203647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,12288,32,0.021632888250880774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,64,0.045435554451412626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,12288,32,0.045444445477591626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,65536,2.1319120195176864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,51200,1.7110409206814234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,65536,4.522097693549262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,16384,0.5487946404351128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,65536,3.912260267469618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,51200,3.5044915941026478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,12288,0.4281662305196126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,16384,1.0153528849283855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,16384,0.9975599712795682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,12288,0.7296426561143664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,10240,0.3543831242455377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,12288,0.7525182300143771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,51200,3.062994639078776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,8192,0.2863653235965305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,10240,0.6043030950758193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,10240,0.6321679751078287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,7168,0.2446533309088813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,8192,0.4784364700317383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,8192,0.5148506694369847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,6144,0.20628533098432753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,7168,0.42253777715894913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,7168,0.47124444113837344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,5120,0.16896355152130127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,6144,0.3601795567406549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,6144,0.39472267362806535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,4096,0.13748621940612793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,5120,0.2964471181233724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,5120,0.3246106571621365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,3584,0.11689066886901855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,4096,0.23406312200758192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,4096,0.262048880259196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,3072,0.10002844201193915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,3584,0.23155556784735787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,3072,0.17857778072357178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,3584,0.20618400308820936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,3072,0.202836447291904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,2560,0.08632444673114353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,2048,0.07180355654822455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,2560,0.15303199821048313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,2560,0.1746284431881375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,1536,0.05335999859703911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,2048,0.12623555130428737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,2048,0.1452746656205919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,1024,0.04235555397139656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,1536,0.1184506681230333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,1536,0.10057155291239421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,1024,0.07558933231565687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,768,0.030977778964572485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,512,0.023333332604832117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,1024,0.09062133232752483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,768,0.07706577910317315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,768,0.061908443768819175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,256,0.017873777283562552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,512,0.050372444921069674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,512,0.0626479983329773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,128,0.017100445098347135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,256,0.049602667490641274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,256,0.04171644316779243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,64,0.014496889379289416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,10240,128,0.04009333252906799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,128,0.03762222329775492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,10240,32,0.015085332923465304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,32,0.03784266776508755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,10240,64,0.03768710957633124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,65536,1.7716542349921331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,51200,1.28875732421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,65536,3.8947287665473094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,65536,3.5213635762532554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,16384,0.45119110743204754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,16384,0.8725066714816623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,51200,3.020899454752604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,12288,0.3233715428246392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,16384,0.8999030854966906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,10240,0.29722489251030815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,12288,0.6360924508836534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,12288,0.6774969100952148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,10240,0.5271226565043131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,8192,0.24240978558858237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,51200,2.762386745876736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,10240,0.5709777937995063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,7168,0.21173777845170763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,8192,0.4210035536024306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,8192,0.46945598390367294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,6144,0.1759413348303901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,7168,0.3698648876614041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,7168,0.4088791211446126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,5120,0.14410666624704996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,6144,0.3169822163052029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,4096,0.1121893326441447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,6144,0.3523964352077908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,5120,0.2605617841084798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,5120,0.29266312387254506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,3584,0.09990133179558648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,4096,0.20471378167470297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,4096,0.23576800028483072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,3072,0.08568977647357517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,3584,0.18079377545250785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,3584,0.20899911721547446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,2560,0.07374399900436401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,3072,0.18199911382463244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,3072,0.1575982173283895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,2048,0.06229155593448215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,2560,0.13433955775366888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,1536,0.04730133215586344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,2048,0.13143732812669542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,2048,0.11163200272454156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,1536,0.08824711375766331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,1024,0.03357955482270982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,1536,0.10662133163876003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,768,0.02706844442420536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,1024,0.06617422236336602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,1024,0.0827546649509006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,768,0.05356088942951626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,512,0.021027555068333943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,768,0.0698488884501987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,256,0.01629066632853614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,512,0.04292355643378364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,512,0.057777775658501514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,128,0.014499555031458536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,256,0.0344142218430837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,256,0.04483555422888862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,2560,0.1568924453523424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,64,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,8192,32,0.013610666824711693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,8192,128,0.03501510951254103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,128,0.0329146683216095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,64,0.03130844566557143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,8192,32,0.0314062237739563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,51200,1.3103271060519748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,65536,1.5527608659532335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,65536,3.488449732462565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,16384,0.41611822446187335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,65536,3.3937430911593967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,51200,2.7507627275254993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,12288,0.3097439871893989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,51200,2.6554959615071616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,16384,0.8143644332885742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,10240,0.2322631147172716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,12288,0.589829338921441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,16384,0.8666790856255425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,8192,0.20743821726904976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,12288,0.6540746688842773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,10240,0.48995643191867405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,8192,0.3941520055135091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,7168,0.16544800334506565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,8192,0.44612354702419704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,6144,0.15042844083574083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,7168,0.3440675470564101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,7168,0.3937555683983697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,6144,0.30027198791503906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,5120,0.12373333507113987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,6144,0.3387004534403483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,10240,0.549183103773329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,5120,0.2423093318939209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,4096,0.09842044115066528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,3584,0.09012533558739556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,4096,0.18981333573659262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,4096,0.2264711062113444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,3072,0.0758097767829895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,3584,0.16679910818735758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,3584,0.2000888850953844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,2560,0.06387555599212646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,3072,0.14640355110168457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,3072,0.1754773325390286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,2560,0.12503555085923937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,2048,0.05006844467586941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,2560,0.15095999505784777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,2048,0.10367111365000407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,2048,0.12585066424475774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,5120,0.28227821985880536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,1536,0.03959644503063626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,1024,0.026893332600593567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,1536,0.08237066533830431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,1024,0.061569776799943715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,768,0.02293066680431366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,1536,0.10377777947319879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,1024,0.07888622416390313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,768,0.04906577865282694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,512,0.017296888762050204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,768,0.06801066795984904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,256,0.01332444449265798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,512,0.039656000004874334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,512,0.05527289046181572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,128,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,256,0.04242222176657783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,256,0.03267822331852383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,128,0.02886311213175456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,64,0.014864888456132678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,7168,128,0.03398133317629496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,7168,32,0.015559110376569959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,32,0.028563555743959215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,65536,1.3568177753024633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,7168,64,0.028799109988742407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,51200,1.0646986431545682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,65536,3.2866221533881297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,16384,0.3382577896118164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,16384,0.7421342002020942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,51200,2.5499369303385415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,65536,3.255371517605252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,12288,0.2573626571231418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,51200,2.5468080308702254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,16384,0.8330951266818576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,10240,0.2323546674516466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,12288,0.5448115666707357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,12288,0.6303288671705458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,10240,0.45351113213433164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,8192,0.17210043801201713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,7168,0.1537217828962538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,10240,0.5257502131991917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,8192,0.36163287692599827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,8192,0.428821325302124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,6144,0.12831021679772273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,7168,0.3776222334967719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,7168,0.31941244337293834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,6144,0.27287289831373424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,5120,0.10733244154188369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,4096,0.08751644690831502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,6144,0.3253973325093587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,5120,0.27014756202697754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,5120,0.22469067573547363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,4096,0.17585955725775826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,3584,0.07902489105860393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,4096,0.2167511118782891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,3072,0.0652373300658332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,3584,0.19184888733757866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,3072,0.168266667260064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,2560,0.056567112604777016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,2560,0.11552977561950684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,2560,0.14395199881659612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,2048,0.04517689016130236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,2048,0.09600444634755452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,2048,0.1210462252298991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,1536,0.034464889102511935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,1536,0.07597244448131986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,1536,0.09873777627944946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,1024,0.024246222443050806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,3584,0.15490044487847224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,3072,0.1337733268737793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,768,0.022156443860795762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,1024,0.07587644126680163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,512,0.018108444081412423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,768,0.044617778725094266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,512,0.03551466597451104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,256,0.013939556148317126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,768,0.0653182201915317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,512,0.0525857773092058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,128,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,256,0.03938400083118015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,128,0.02645422187116411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,64,0.011535999675591787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,6144,128,0.033512890338897705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,1024,0.056442664729224316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,6144,32,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,64,0.026192888617515564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,32,0.026111111044883728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,6144,256,0.028910222980711196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,65536,1.1881528430514865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,51200,0.801163567437066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,65536,2.9711956448025174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,16384,0.2790577676561144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,51200,2.2811830308702254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,65536,2.9927315182156033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,16384,0.6752239863077799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,51200,2.3471893734402127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,12288,0.2090675565931532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,16384,0.7682399749755859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,10240,0.18062310748630098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,12288,0.499525335099962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,12288,0.5806346469455296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,8192,0.14537599351671007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,10240,0.41627732912699383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,10240,0.48952886793348527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,7168,0.12595555517408583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,8192,0.3327431148952908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,8192,0.396250671810574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,7168,0.29313956366644967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,6144,0.11082488960689968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,7168,0.3494444423251682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,6144,0.25124533971150714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,5120,0.0941271119647556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,6144,0.3006275494893392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,4096,0.0763182242711385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,5120,0.20592178238762748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,5120,0.2492755519019233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,3584,0.0655848913722568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,4096,0.1621777746412489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,4096,0.2005866633521186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,3072,0.058506667613983154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,3584,0.14178400569491917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,3584,0.17795911100175646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,3072,0.12302933798895942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,2560,0.04741955465740628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,3072,0.15561778015560573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,2048,0.03983022107018365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,2560,0.10496355427636041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,2560,0.13412888844807944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,1536,0.030003554291195337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,2048,0.08845866388744777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,2048,0.11237866348690456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,1024,0.02321333355373806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,1536,0.07006933291753133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,1536,0.09149510992897882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,1024,0.05070933368470934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,768,0.018180444836616516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,1024,0.07059999969270495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,768,0.04100355505943298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,768,0.05995555718739828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,512,0.013190222283204397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,256,0.010566222170988718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,512,0.032079110542933144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,128,0.009571555587980483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,256,0.037688887781567044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,256,0.02681244413057963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,64,0.009053332938088311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,128,0.024029332730505202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,128,0.031231111950344507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,5120,32,0.009355555805895064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,64,0.024051555328898962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,65536,0.8198693593343099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,5120,512,0.047945777575174965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,5120,32,0.02365511159102122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,51200,0.6882088979085287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,65536,2.6178275214301214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,16384,0.23456711239284941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,51200,2.0489039950900607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,16384,0.6385768784417046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,12288,0.17086399926079643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,65536,2.8651031917995877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,51200,2.242653316921658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,16384,0.7336951361762153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,10240,0.15322666698031956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,12288,0.4554328918457031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,12288,0.5561493237813314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,10240,0.3793697886996799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,8192,0.11902756161159939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,7168,0.1065831118159824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,10240,0.46721246507432723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,8192,0.3044559955596924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,8192,0.3783875571356879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,6144,0.0884702205657959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,7168,0.2682604524824354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,7168,0.3342675632900662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,5120,0.07700800233417086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,6144,0.2293297714657254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,6144,0.28819200727674693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,4096,0.05851288636525472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,5120,0.18845600552029082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,3584,0.055162668228149414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,5120,0.2382906542883979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,4096,0.14767377906375462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,4096,0.1908773316277398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,3072,0.04646844334072537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,3584,0.1290231148401896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,3584,0.16946400536431205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,2560,0.040991110934151545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,3072,0.1480462286207411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,2048,0.03282577792803446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,2560,0.09597600168652004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,2560,0.127384000354343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,2048,0.07776800129148695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,1536,0.026318222284317017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,2048,0.10741244422064887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,1536,0.06309955649905734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,1536,0.08730933401319717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,1024,0.018740443719757926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,1024,0.04576177729500664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,1024,0.06708977619806926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,768,0.015821332732836407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,768,0.037885334756639265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,512,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,768,0.05712178018358019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,512,0.03001777662171258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,512,0.045351998673544995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,256,0.009325332939624786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,256,0.035975999302334256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,128,0.008284444610277811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,128,0.021670222282409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,4096,128,0.030348443322711523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,64,0.008972444468074376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,3072,0.11183644665612115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,4096,32,0.009305777649084726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,64,0.021335111724005804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,32,0.021309332715140447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,4096,256,0.02367377777894338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,65536,0.8161999914381238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,51200,0.6677511003282336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,65536,2.4627581702338324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,16384,0.22065777248806426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,51200,2.001729753282335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,65536,2.7489840189615884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,12288,0.17351467079586455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,16384,0.5908400217692057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,12288,0.43233421113755965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,16384,0.7032346725463867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,10240,0.13997511068979898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,12288,0.5322453180948893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,8192,0.1128880050447252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,10240,0.3610880109998915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,51200,2.1417901780870223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,10240,0.44693957434760195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,7168,0.09715021981133355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,8192,0.29025599691602916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,8192,0.3626702096727159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,6144,0.08225155538982816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,7168,0.254838228225708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,7168,0.31985243161519367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,5120,0.0710444450378418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,6144,0.21840622689988878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,4096,0.0563653310139974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,6144,0.27514399422539604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,5120,0.17927822801801893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,5120,0.22668533855014375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,3584,0.049364444282319814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,4096,0.14073333475324842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,4096,0.1834097835752699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,3072,0.04411200020048353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,3584,0.12317244211832683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,2560,0.03468888998031616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,3072,0.14155556095971003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,3072,0.10739199982749091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,2560,0.09157066875033909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,2048,0.02886311213175456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,2560,0.12230311499701606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,1536,0.022671999202834234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,2048,0.07368799712922838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,2048,0.102966222498152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,1024,0.015761777758598328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,1536,0.05960888995064629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,1536,0.08382222387525771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,768,0.012720000412729053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,1024,0.06417600313822429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,1024,0.043656001488367714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,512,0.010449777874681683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,768,0.0352862212393019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,3584,0.16160178184509277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,256,0.00830577810605367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,768,0.055381334490246244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,512,0.04313244422276815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,128,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,256,0.02232088810867733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,256,0.03488622109095255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,512,0.028154667880800035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,64,0.0069395556218094295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3584,32,0.007256888680987888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,128,0.02030044462945726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3584,128,0.028852442900339764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,32,0.02033511135313246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3584,64,0.020636444290479023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,65536,0.631806214650472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,51200,0.5292666753133138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,65536,2.2437307569715714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,16384,0.18328355418311226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,51200,1.793456819322374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,16384,0.5435359742906358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,65536,2.7302381727430554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,51200,2.1384542253282333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,12288,0.1252817842695448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,16384,0.7006613413492838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,10240,0.10711999734242757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,12288,0.4086551136440701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,10240,0.34278400739034015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,8192,0.0858844452434116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,12288,0.5317111015319824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,10240,0.44621864954630536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,7168,0.07867288589477539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,8192,0.27517954508463544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,8192,0.36245155334472656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,6144,0.0645680030186971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,7168,0.3186497688293457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,6144,0.27416443824768066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,6144,0.20741777949863008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,5120,0.05721866422229343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,5120,0.22684976789686415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,4096,0.045137776268853076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,7168,0.24215377701653373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,5120,0.17052000098758271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,3584,0.038704001241260104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,4096,0.13309866852230495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,4096,0.1819626622729831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,3072,0.03417333298259311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,3584,0.11774755848778619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,3584,0.16097244951460096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,2560,0.030728889836205378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,3072,0.101383109887441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,3072,0.14158756203121609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,2560,0.08622488710615371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,2048,0.0244835548930698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,2560,0.12137067317962646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,1536,0.019740444090631273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,2048,0.06991555293401082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,2048,0.10222488641738892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,1536,0.053451554642783276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,1024,0.01457244488928053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,1536,0.0825786656803555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,768,0.011893333660231696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,1024,0.04012889001104567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,1024,0.06407999992370605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,512,0.009140444298585257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,768,0.03327288892534044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,768,0.05297333333227369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,256,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,512,0.0425600012143453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,512,0.02682133350107405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,256,0.021028444170951843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,128,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,256,0.03292977809906006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,64,0.006266666783226862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,128,0.019307555423842538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,3072,32,0.006383111079533895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,3072,128,0.02850488821665446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,64,0.01864355636967553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,3072,32,0.01924622224436866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,65536,0.5592568715413412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,51200,0.44734668731689453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,65536,2.1025049421522355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,16384,0.1431822246975369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,51200,1.6295981936984594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,16384,0.51474486456977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,65536,2.5971031188964844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,12288,0.11110399829016791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,12288,0.38664711846245664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,16384,0.66939910252889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,51200,2.034838146633572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,10240,0.08684621916876899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,12288,0.5068728658888075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,8192,0.07327999671300252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,10240,0.32496444384257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,8192,0.2614702118767632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,7168,0.06315022044711642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,10240,0.42662665579054093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,8192,0.345778677198622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,6144,0.05626489056481255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,7168,0.2297217845916748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,6144,0.1955333285861545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,5120,0.048728889889187284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,7168,0.30542132589552135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,6144,0.2632266680399577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,4096,0.039354665411843195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,5120,0.16050044695536295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,5120,0.2159902254740397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,3584,0.03638489047686259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,4096,0.12552800443437365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,3072,0.031621333625581526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,4096,0.1739493343565199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,3584,0.1102995541360643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,3584,0.15342133575015598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,2560,0.027958220905727808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,3072,0.09565422270033096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,3072,0.13459732797410753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,2048,0.022463111413849726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,2560,0.08088799979951647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,2560,0.11585956149631077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,1536,0.01851733360025618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,2048,0.09755288892322117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,2048,0.06551377640830146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,1536,0.07764177852206759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,1536,0.05061689019203186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,1024,0.013404444687896304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,768,0.01109777804878023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,1024,0.061089780595567494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,768,0.031054221921496924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,1024,0.037787556648254395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,768,0.050422221422195435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,512,0.04105510976579454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,512,0.02474311159716712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,256,0.007002666592597961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,256,0.019720888800091214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,256,0.03216800093650818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,128,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,128,0.017887999614079792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2560,128,0.027151111099455092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,64,0.007015110717879401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,32,0.007283555964628856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,64,0.017589333984586928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2560,32,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2560,512,0.008989333278603023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,65536,0.4436515437232123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,51200,0.36467377344767254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,65536,1.925679948594835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,51200,1.5112667083740234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,16384,0.11423644754621719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,65536,2.4664861891004777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,16384,0.483932442135281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,12288,0.08900355630450779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,16384,0.6356319851345485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,12288,0.36535554462009007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,10240,0.07396355602476332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,51200,1.9334061940511067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,12288,0.48200352986653644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,8192,0.06418577829996745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,10240,0.306821346282959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,8192,0.24663911925421822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,10240,0.40586222542656797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,7168,0.05392000079154968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,8192,0.3300408787197537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,6144,0.04616711205906338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,7168,0.21643822722964814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,7168,0.29043732749091256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,5120,0.03994133406215244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,6144,0.1856400039460924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,6144,0.24926310115390352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,4096,0.03219644559754266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,5120,0.15230045053693983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,5120,0.20677688386705187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,3584,0.028406222661336262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,4096,0.11868888801998562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,4096,0.16447999742296007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,3072,0.024007999234729346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,3584,0.10363466872109307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,3584,0.1461368931664361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,2560,0.021388444635603163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,3072,0.08969066540400188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,3072,0.12834044297536215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,2048,0.01814844376511044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,2560,0.076109336482154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,2560,0.11064355240927802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,1536,0.014279110564125909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,2048,0.06298844681845771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,2048,0.09258755710389878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,1024,0.010695111420419482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,1536,0.04832977718777127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,1536,0.0759262243906657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,768,0.009190222455395592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,1024,0.03496888942188687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,1024,0.0576160020298428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,512,0.0069955554273393415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,768,0.04795111219088236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,768,0.029297779003779095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,512,0.023341332872708637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,256,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,128,0.00562755556570159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,512,0.04011288947529263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,256,0.018615111708641052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,256,0.0314062237739563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,64,0.0052284445199701525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,2048,32,0.0052560000783867305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,2048,128,0.02606044378545549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,128,0.017284444636768766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,64,0.016495111915800307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,2048,32,0.016508445143699646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,65536,0.33864532576666934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,51200,0.26412977112664116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,51200,1.4149253633287218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,65536,1.8125822279188368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,16384,0.08829333384831746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,16384,0.4572230974833171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,51200,1.9297235276963975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,12288,0.06477066543367174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,65536,2.4623635609944663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,10240,0.05613422393798828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,12288,0.34628799226548934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,10240,0.2910062207116021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,12288,0.48333602481418186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,8192,0.046281778150134616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,10240,0.40581244892544216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,7168,0.04083733426200019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,8192,0.23478221893310547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,8192,0.3287048869662815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,6144,0.03651644455062018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,16384,0.6343910959031847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,7168,0.2057804399066501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,5120,0.02931200133429633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,7168,0.2908053398132324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,6144,0.2495200104183621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,5120,0.1437528928120931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,5120,0.20519910918341744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,4096,0.024692444337738886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,4096,0.11198666360643174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,3584,0.022808000445365906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,4096,0.16481599542829725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,3584,0.09856800238291423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,3072,0.020371556282043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,6144,0.175982223616706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,3584,0.14591199821896023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,2560,0.017514665921529133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,3072,0.08619911140865749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,3072,0.12734044922722712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,2048,0.014367111855083041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,2560,0.07311644156773885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,2560,0.10975555578867595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,1536,0.012484444512261285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,2048,0.05994844436645508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,2048,0.09269599782096015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,1536,0.04626577761438158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,1536,0.07477333148320515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,1024,0.009396444592210982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,768,0.008238222036096785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,1024,0.056503110461764865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,768,0.028023110495673284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,768,0.047932446002960205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,512,0.007204444044166141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,512,0.02198311189810435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,256,0.0059262220230367445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,512,0.038689778910742864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,256,0.017549332645204332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,256,0.030454221698972914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,128,0.004902222090297275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1536,128,0.02541333271397485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,64,0.004578666554556953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,128,0.015879111157523263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,1024,0.032292445500691734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1536,32,0.004599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,64,0.015466666883892484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1536,32,0.015212444795502557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,65536,0.23573332362704805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,51200,0.1776284509234958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,65536,1.6769012875027125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,51200,1.315633773803711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,16384,0.06217866473727756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,16384,0.4285128911336263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,12288,0.04694044589996338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,16384,0.6024382379319932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,65536,2.3344559139675565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,10240,0.039107554488711886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,12288,0.32510577307807076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,12288,0.457931571536594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,10240,0.2721937762366401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,8192,0.03287911083963182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,8192,0.21995200051201713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,10240,0.38598044713338214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,7168,0.02868622210290697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,8192,0.3126728799608019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,51200,1.8303146362304688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,6144,0.02462844385041131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,7168,0.19309955173068574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,7168,0.2762835555606418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,5120,0.021145777569876775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,6144,0.23727999793158638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,6144,0.16532799932691786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,4096,0.01774488886197408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,5120,0.1342044406467014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,3584,0.015653333730167814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,4096,0.1048133307033115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,5120,0.19519821802775064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,4096,0.1563431157006158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,3072,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,3584,0.0916551086637709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,2560,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,3072,0.08005955484178331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,3584,0.13893689049614802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,2048,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,3072,0.12107022603352864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,2560,0.06758666700787015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,2560,0.10503200027677749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,1536,0.008973333570692275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,2048,0.0544088880221049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,2048,0.08797244230906169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,1024,0.007306666837798224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,1536,0.04149422380659316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,1536,0.07065511412090726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,768,0.006568888823191325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,1024,0.05266844564014011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,1024,0.030485332012176514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,512,0.005621333503060871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,768,0.025928000609079998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,768,0.04550666610399882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,256,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,512,0.0207360006040997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,256,0.016201777590645682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,512,0.03707289033465915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,256,0.029498666524887085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,128,0.0042071110672420925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,64,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,1024,128,0.02444711161984338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,1024,32,0.004084444294373195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,64,0.013785777820481194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,32,0.013939556148317126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,65536,0.20202933417426217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,1024,128,0.01442399952146742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,51200,0.15669243865542942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,65536,1.632905748155382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,65536,2.3348276350233292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,51200,1.277740478515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,16384,0.0535715553495619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,16384,0.4173368877834744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,12288,0.0414959987004598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,16384,0.6025004386901855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,12288,0.3161991172366672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,51200,1.830352783203125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,10240,0.034224000242021345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,12288,0.4575040075514052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,8192,0.02786311176088121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,10240,0.2658640013800727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,10240,0.38495734002855087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,7168,0.02422399984465705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,8192,0.21395822366078696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,8192,0.31276978386773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,6144,0.021539555655585393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,7168,0.27558665805392796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,6144,0.1605964501698812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,5120,0.018607111440764535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,6144,0.23632711834377715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,5120,0.13089866108364528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,4096,0.015503999259736804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,5120,0.19450399610731336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,4096,0.10264088710149129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,4096,0.1559013393190172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,3584,0.013896889156765409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,7168,0.18840355343288848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,3072,0.012356444365448423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,3584,0.08922488821877374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,3584,0.13897866672939724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,2560,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,3072,0.07826755444208781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,3072,0.12088178263770209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,2048,0.00942666663063897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,2560,0.06584178076850043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,2560,0.10464621914757623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,1536,0.00832799987660514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,2048,0.052258667018678456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,2048,0.08695911036597358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,1024,0.0069617778062820435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,1536,0.03997244437535604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,1536,0.06862044334411621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,768,0.005976000179847081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,1024,0.029531555043326482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,1024,0.0520631108019087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,512,0.004938666605287128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,768,0.025176889366573755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,768,0.044970668024486966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,512,0.01997422178586324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,512,0.03631644447644552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,256,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,256,0.028811554114023846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,128,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,768,128,0.024255999260478552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,64,0.0038559999730851916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,64,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,32,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,32,0.013423110875818463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,768,256,0.004845333182149463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,65536,0.1445173290040758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,768,128,0.013762666947311826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,51200,0.11634222666422527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,65536,1.5556231604682074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,16384,0.040082666609022356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,51200,1.2185217539469402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,16384,0.39945777257283527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,12288,0.029363556040657893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,65536,2.2027528550889754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,16384,0.5694168938530816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,10240,0.023619555764728125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,12288,0.43257776896158856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,10240,0.2544328901502821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,8192,0.020398222737842135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,10240,0.36390667491488987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,51200,1.729162640041775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,7168,0.01663466625743442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,8192,0.20528711213005912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,8192,0.29484711752997506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,6144,0.015062222878138224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,7168,0.2600933445824517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,7168,0.18043111430274117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,12288,0.3030453258090549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,5120,0.012727999852763282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,6144,0.15401511722140843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,4096,0.011023999916182624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,5120,0.12538666195339626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,5120,0.18376266956329346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,4096,0.09795822037590875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,3584,0.010117332968446944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,4096,0.14718133873409694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,3584,0.08526221911112468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,3072,0.008990222381220924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,3584,0.13042133384280735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,3072,0.07415022452672322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,2560,0.007989333735571967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,3072,0.11407821708255345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,2560,0.06256444586647882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,6144,0.22346043586730957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,2048,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,2560,0.09883289204703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,1536,0.006648889018429651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,2048,0.04934222168392605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,2048,0.08225066794289483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,1024,0.00527555536892679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,1536,0.03827111257447137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,1536,0.06659911076227824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,768,0.004888888862397936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,1024,0.028018666638268366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,1024,0.04970844586690267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,512,0.004295999805132548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,768,0.04295644495222303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,768,0.024215110474162634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,256,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,512,0.035233779086007014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,512,0.019319110446506076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,256,0.014859555496109856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,128,0.003604444364706675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,256,0.02767733401722378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,64,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,128,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,512,32,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,512,128,0.023997333314683702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,64,0.012869333227475485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,65536,0.09640444649590386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,512,32,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,51200,0.07620177666346233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,51200,1.1857804192437065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,65536,1.5149368709988063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,16384,0.031231999397277832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,16384,0.3896435631646051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,65536,2.1942861345079208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,12288,0.022848889231681824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,16384,0.5663688977559408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,12288,0.29540710979037815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,10240,0.018965333700180054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,51200,1.7168888515896266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,8192,0.015253333581818474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,12288,0.4305493301815457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,10240,0.24818756845262316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,10240,0.3629306687249078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,8192,0.19955378108554414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,8192,0.2934044467078315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,6144,0.01238844460911221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,7168,0.1761253277460734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,7168,0.2587706777784559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,6144,0.14984177218543157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,5120,0.010683555569913654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,6144,0.222433778974745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,5120,0.12212088372972275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,5120,0.18194222450256348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,4096,0.008917333351241218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,3584,0.008267555799749162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,4096,0.09517511394288804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,4096,0.1458373334672716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,3072,0.007336888876226213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,7168,0.013953778478834363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,3584,0.08323199881447686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,2560,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,3584,0.12922578387790257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,3072,0.07133599784639147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,3072,0.11292710569169785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,2048,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,2560,0.09740977817111546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,2560,0.06085955434375339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,1536,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,2048,0.04765244325002035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,2048,0.08100533485412598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,1024,0.004896000027656555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,1536,0.0639902220831977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,1536,0.037350220812691584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,1024,0.02716800073782603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,768,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,1024,0.04898755417929756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,512,0.0038488888078265717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,768,0.04251111216015286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,768,0.023373333944214716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,512,0.03490044342146979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,256,0.0034906665484110513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,128,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,256,0.02747822304566701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,256,0.014459555347760519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,128,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,64,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,256,32,0.0032604444358083936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,256,128,0.023410666320059035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,64,0.012113778127564324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,65536,0.07730488644705878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,32,0.01239822225438224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,51200,0.061807996696896024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,256,512,0.01860000027550591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,65536,1.4768977695041234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,51200,1.1589004728529189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,16384,0.02499733368555705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,65536,2.1912648942735458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,16384,0.37964532110426163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,12288,0.01701244380739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,16384,0.5661635398864746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,12288,0.2880551020304362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,10240,0.014888889259762235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,51200,1.7166097429063585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,12288,0.429733329349094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,8192,0.012038222617573209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,10240,0.2420577738020155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,10240,0.36166045400831437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,7168,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,8192,0.19509689013163248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,8192,0.29371645715501576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,7168,0.17131733894348145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,7168,0.2584737671746148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,6144,0.1458453337351481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,5120,0.008643555144468943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,6144,0.22259555922614205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,5120,0.11903288629319932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,5120,0.18197333812713623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,4096,0.007647111184067196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,4096,0.09244977765613133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,4096,0.14548356003231472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,3584,0.006931555353932911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,6144,0.010057777994208867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,3072,0.006264888991912206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,3584,0.08034933275646634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,3584,0.12904622819688585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,2560,0.006266666783226862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,3072,0.11221422089470758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,2560,0.058607108063168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,2048,0.00573244442542394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,3072,0.06964266962475248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,2560,0.09730755620532566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,2048,0.045936889118618436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,1536,0.005111999809741974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,1024,0.00444177786509196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,2048,0.0808417797088623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,1536,0.03577244281768799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,768,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,1536,0.06405422422620985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,1024,0.048546665244632296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,768,0.022890667120615642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,768,0.0422364440229204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,512,0.003955555458863576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,256,0.003612444632583194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,1024,0.02677511175473531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,512,0.03424355718824599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,128,0.0031724443866146933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,256,0.013818666338920593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,256,0.02718488872051239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,64,0.0029288888391521242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,128,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,128,32,0.003148444410827425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3328,128,128,0.02233155568440755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,65536,0.07304622067345513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,64,0.012149333126015134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,32,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,51200,0.059699555238087974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,16384,0.024919999970330134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,128,512,0.017856889300876193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,12288,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,16384,0.37951821751064724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,10240,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,12288,0.28739910655551487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,51200,1.1571377648247612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,8192,0.01219733307758967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,7168,0.011047999891969891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,10240,0.24187911881340873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,65536,1.4788249333699544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,6144,0.009684444301658207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,8192,0.19497866100735137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,5120,0.008647110727098253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,7168,0.17096622784932455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,4096,0.007291555404663086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,3584,0.006931555353932911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,6144,0.14668533537122938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,5120,0.11875111526913112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,3072,0.00658044425977601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,2560,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,4096,0.09166933430565728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,2048,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,3584,0.0803297758102417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,2560,0.05794755617777506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,1536,0.004909333255555895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,2048,0.04558666547139486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,1024,0.004280888785918553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,768,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,1536,0.034690668185551964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,512,0.003536888708670934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,1024,0.026094221406512793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,256,0.003245333416594399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,768,0.022708444131745234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,512,0.01789422167672051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,128,0.0031395554542541504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,64,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,256,0.013462222284740873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,128,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,64,32,0.0031591111587153543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,64,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,65536,0.0746648907661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,32,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,51200,0.0581173300743103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,64,3072,0.06942222515741985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,16384,0.022039999564488728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,16384,0.3790648778279622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,12288,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,12288,0.28755911191304523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,10240,0.010319111247857412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,51200,1.1578630871242948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,8192,0.014471999473041959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,65536,1.4763413535224066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,7168,0.009088888764381409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,10240,0.24143645498487684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,6144,0.008685333033402761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,8192,0.19505866368611655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,5120,0.010049777726332346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,7168,0.1711031066046821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,4096,0.008980444735950893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,6144,0.1463493373658922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,3584,0.008288000192907121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,5120,0.11866488721635605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,3072,0.007280000381999546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,4096,0.09183288945092095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,2560,0.006594666590293248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,3584,0.08049333095550537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,2048,0.004937777916590373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,3072,0.06951910919613309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,1536,0.004534222185611725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,2560,0.05815644396675957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,1024,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,2048,0.04578311244646708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,768,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,1536,0.03559199968973795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,512,0.0032631111227803757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,1024,0.02604266670015123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,256,0.003157333367400699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,768,0.02258133391539256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,128,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,512,0.01789955629242791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,64,0.0028320000403457214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,256,0.013398222625255585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3328,32,32,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,128,0.01202755586968528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,64,0.01146044416560067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3328,32,32,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,16384,3.0792694091796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,12288,2.3232409159342446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,16384,4.51454120212131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,12288,3.1074061923556857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,10240,1.9484142727322047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,51200,9.773734198676214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,8192,1.586186620924208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,10240,2.5397830539279513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,8192,1.952743106418186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,7168,1.383801778157552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,6144,1.1851022508409288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,7168,1.7078044679429796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,6144,1.4554328918457031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,5120,1.2013538148668077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,4096,0.7931520144144694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,5120,1.0194213655259874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,51200,15.483111911349825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,4096,0.9475662443372938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,3584,0.731694221496582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,3072,0.6514817873636881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,3584,0.8339893552992078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,2560,0.5340524249606663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,3072,0.7152373525831434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,2048,0.4168684482574463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,1536,0.32975999514261883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,2048,0.49795113669501406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,2560,0.6279084417555068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,1024,0.23033777872721353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,768,0.17310310734642875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,1536,0.39333332909478086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,512,0.12983466519249812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,768,0.24670044581095377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,256,0.0977591143714057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,512,0.20108177926805285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,128,0.07967999908659193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,1024,0.2981208960215251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,64,0.06734933455785115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,256,0.17848177750905356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,65536,32,0.07017333639992608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,128,0.1697404384613037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,32,0.17550222078959146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,65536,64,0.17331555154588488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,51200,7.563276502821181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,65536,9.941533406575521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,16384,2.5032471550835504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,12288,1.8578133053249781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,51200,12.335289001464844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,16384,3.572289784749349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,10240,1.5670773188273113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,12288,2.476867463853624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,65536,15.553697374131945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,8192,1.238443586561415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,10240,2.022475560506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,7168,1.085304896036784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,7168,1.3682649400499132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,8192,1.562154663933648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,6144,0.9652737511528863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,5120,0.7872533268398709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,6144,1.1637626224093967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,4096,0.6491111119588217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,5120,0.9620559480455187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,4096,0.7587599754333496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,3584,0.5663484467400445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,3072,0.5002044571770562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,3072,0.5743155479431152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,2560,0.40823109944661456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,2048,0.3258693218231201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,2560,0.4888728989495172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,1536,0.26812267303466797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,3584,0.6683182186550565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,2048,0.39958932664659286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,1024,0.18924000528123644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,768,0.15106399854024252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,1536,0.31515823470221627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,512,0.11284266577826606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,768,0.19123555554283989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,1024,0.23900532722473145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,256,0.07992177539401583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,512,0.16133866045210096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,128,0.06292266978157891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,64,0.0544088880221049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,256,0.143251551522149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,51200,32,0.05773422453138563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,128,0.13582221666971842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,32,0.14064710670047337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,51200,64,0.13821510473887125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,51200,2.349365234375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,65536,3.133094151814779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,16384,0.7640257941351997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,12288,0.5602729055616591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,16384,1.3203129238552518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,51200,4.626162634955512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,10240,0.5263857841491699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,12288,0.9488506317138672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,8192,0.45015642378065324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,8192,0.6164337793986002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,10240,0.782918241288927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,65536,5.932854970296224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,7168,0.40926400820414227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,6144,0.32139465543958873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,5120,0.2793360021379259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,7168,0.5404462284511989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,4096,0.22861777411566842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,6144,0.46170488993326825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,3584,0.1929466724395752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,5120,0.3867502212524414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,3072,0.1644373337427775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,3584,0.26257777214050293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,2560,0.14050666491190592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,3072,0.22855999734666613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,2048,0.11258399486541748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,2560,0.1951155530081855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,1536,0.08827377690209283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,1024,0.062309331364101835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,1536,0.127292447619968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,2048,0.1611893309487237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,768,0.048102223210864596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,512,0.037258668078316584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,768,0.07766577932569715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,512,0.06562311119503446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,4096,0.3006239997016059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,128,0.02268444498380025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,256,0.056681778695848256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,64,0.01944266590807173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,128,0.053131557173199125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,32,0.01996977792845832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,64,0.053545776340696544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,32,0.05433155430687798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,16384,1024,0.09653955698013306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,16384,256,0.02686133318477207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,51200,2.0265030331081815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,65536,2.4228515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,16384,0.6559679773118761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,16384,1.065640025668674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,12288,0.46400176154242617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,10240,0.40348267555236816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,12288,0.7700764338175455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,51200,3.7305530971950955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,10240,0.6391839981079102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,8192,0.32905599806043834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,65536,4.704385969373915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,7168,0.2991093264685737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,8192,0.5061555438571507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,6144,0.2582320107354058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,5120,0.20654133955637613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,7168,0.44375377231174046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,6144,0.3805928760104709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,5120,0.3113120132022434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,3584,0.14174666669633654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,4096,0.24561688635084364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,3072,0.12434666686587864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,3584,0.21586399608188203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,2560,0.10113333331214057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,3072,0.18696355819702148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,2048,0.08337244722578262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,2560,0.16075021690792507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,1536,0.06478133466508654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,2048,0.13234755727979872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,4096,0.16698133945465088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,1024,0.04805333415667216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,768,0.03782577647103204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,1536,0.10510133372412787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,512,0.029472000069088403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,768,0.06397244665357801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,1024,0.08030133114920722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,256,0.022064889470736187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,512,0.05446666810247633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,128,0.01864355636967553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,64,0.015903110305468243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,128,0.042450666427612305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,256,0.04639377858903673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,12288,32,0.016174222032229107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,64,0.0425146652592553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,12288,32,0.042507555749681264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,51200,1.556504037645128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,65536,1.9498888651529949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,16384,0.5102631251017252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,16384,0.9441021813286675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,12288,0.37893422444661456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,65536,4.22777345445421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,10240,0.30713510513305664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,12288,0.682430214352078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,8192,0.25221067004733616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,7168,0.21895289421081543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,8192,0.4504391352335612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,51200,3.264919916788737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,7168,0.3950168821546767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,6144,0.20253154966566297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,5120,0.16421600182851157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,6144,0.3361368974049886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,4096,0.13152978155348036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,5120,0.2760159969329834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,3584,0.1172320048014323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,3072,0.10257777902815078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,10240,0.5666400061713325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,4096,0.21893510553571913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,3584,0.19204089376661512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,2560,0.0854533314704895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,2048,0.07109866539637248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,3072,0.16709600554572213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,1536,0.054894222153557666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,1024,0.04007200068897671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,2048,0.11810844474368626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,2560,0.14427288373311362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,768,0.03008355696996053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,512,0.022056889202859666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,1536,0.09425866603851318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,256,0.01789155602455139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,1024,0.07118044296900432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,768,0.05771733654869927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,128,0.01517777807182736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,512,0.047873778475655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,64,0.012825777961148156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,256,0.03939911060863071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,10240,32,0.013166222307417126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,128,0.036299556493759155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,64,0.03463911016782125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,10240,32,0.034656888908810086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,51200,1.1925680372450087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,65536,1.518471082051595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,16384,0.419814215766059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,12288,0.3066115644243029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,16384,0.8143848843044705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,10240,0.2755608823564318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,12288,0.6054764323764378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,51200,2.813235600789388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,8192,0.24091821246676973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,10240,0.5030115445454916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,7168,0.2139208846622043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,6144,0.1739208830727471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,8192,0.3939813243018256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,65536,3.559551239013672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,5120,0.1390568945142958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,7168,0.34586310386657715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,4096,0.11111644903818767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,6144,0.3099822203318278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,3584,0.0993573334481981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,4096,0.19003378020392525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,5120,0.24252798822191027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,3072,0.08591466479831272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,3584,0.16832710636986625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,2048,0.05968177980846829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,3072,0.14658310678270128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,2560,0.1259875562455919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,1536,0.047163556019465126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,1024,0.03439822130733066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,2048,0.10423111253314549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,768,0.027487110760476854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,1536,0.08312266402774386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,512,0.019306666321224637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,1024,0.061978664663102895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,768,0.049350221951802574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,256,0.015856888559129503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,512,0.040729777680502996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,128,0.013035555680592855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,256,0.03217333224084642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,64,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,128,0.03081688947147793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,2560,0.07354755534066094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,8192,32,0.011468444433477191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,64,0.029888000753190782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,8192,32,0.029935999049080744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,51200,1.1266969045003254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,65536,1.4571946461995442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,16384,0.3870053291320801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,16384,0.7854195700751411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,12288,0.2846862210167779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,65536,3.354208840264214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,12288,0.5532106823391384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,10240,0.25923911730448407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,8192,0.2014791170756022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,51200,2.5791138543023004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,10240,0.45801422331068253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,8192,0.3873324394226074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,6144,0.15111645062764487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,7168,0.32180711958143443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,5120,0.1196017795138889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,4096,0.10138489140404595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,5120,0.2248284551832411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,3584,0.08504444360733032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,4096,0.17644177542792427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,7168,0.1703102191289266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,3072,0.07717421982023451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,2560,0.06115466356277466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,3584,0.15524178081088597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,2048,0.05502577622731527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,3072,0.13696800337897405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,2560,0.1175653272204929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,1536,0.03837777839766608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,1024,0.02958844436539544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,2048,0.09713421927558051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,768,0.022848000129063923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,1536,0.07726222276687622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,512,0.018017777138286166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,6144,0.273796452416314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,256,0.013154666456911298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,1024,0.05755911270777384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,768,0.045816888411839805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,512,0.03712177938885159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,128,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,256,0.029502222935358684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,32,0.010048888623714447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,128,0.027859555350409612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,64,0.02792177928818597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,7168,32,0.027445332871543035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,51200,0.9386311637030708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,7168,64,0.010018666585286459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,65536,1.189151128133138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,16384,0.30012533399793834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,12288,0.244558228386773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,16384,0.6909333335028754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,10240,0.2032488849427965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,12288,0.5299022462632921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,8192,0.17078044679429796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,10240,0.43914400206671816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,7168,0.1467262241575453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,8192,0.3381200101640489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,65536,2.999120924207899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,6144,0.1277946631113688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,7168,0.29736354615953237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,5120,0.1034053365389506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,4096,0.08445333109961616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,51200,2.367892371283637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,5120,0.2074373298221164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,3584,0.07319911321004231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,4096,0.16377777523464626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,3072,0.06473955843183729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,2560,0.05566933419969347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,3584,0.14388622177971735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,2048,0.04558577802446154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,3072,0.12553866704305014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,1536,0.0355404449833764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,2560,0.10849244727028741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,1024,0.024396444360415142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,2048,0.08951377868652344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,768,0.019654222660594516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,1536,0.07098399930530123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,512,0.015536889433860779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,1024,0.05216888917817009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,256,0.011702222128709158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,768,0.04221955604023404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,512,0.03445511062939962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,128,0.010088000032636855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,256,0.027148443791601393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,64,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,6144,0.25321867730882436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,6144,32,0.00963288876745436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,128,0.02575911084810893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,64,0.025423111187087163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,6144,32,0.025511110822359722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,51200,0.7746515803866916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,65536,0.9797759585910373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,16384,0.26524355676439076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,16384,0.6568391058180068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,12288,0.20340089003245035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,65536,2.6905494266086154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,12288,0.46667024824354386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,10240,0.15840000576443142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,51200,2.1538914574517145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,8192,0.13887288835313585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,10240,0.38860177993774414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,7168,0.11878044075436062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,6144,0.10149421956804063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,8192,0.31087732315063477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,5120,0.08333510822719997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,7168,0.27337421311272514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,4096,0.07059644328223334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,6144,0.23648532231648764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,3584,0.06074755721622043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,5120,0.19024711185031465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,4096,0.1496853298611111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,3584,0.13203110959794787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,2560,0.045243554645114474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,3072,0.11447200510236953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,2048,0.038155555725097656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,2560,0.09840444723765056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,1536,0.02961955467859904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,2048,0.0820488863521152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,1024,0.022005332840813532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,1536,0.06516977813508776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,768,0.01773066653145684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,1024,0.04704888992839389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,512,0.013024888932704926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,768,0.038776887787712946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,256,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,3072,0.05539910991986593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,128,0.008618666893906063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,256,0.025088888075616624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,128,0.023223110371165808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,32,0.007956444389290279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,5120,64,0.008172444171375698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,64,0.023370666636361018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,32,0.023061333431137934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,65536,0.8107591205173068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,51200,0.6427386601765951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,5120,512,0.030665778451495703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,16384,0.23011199633280435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,12288,0.1729280021455553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,16384,0.5704906781514486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,10240,0.14117956161499023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,12288,0.423509332868788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,8192,0.1138497723473443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,10240,0.35383023156060117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,7168,0.09945866796705459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,8192,0.2840569019317627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,65536,2.399579577975803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,6144,0.08723022540410359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,5120,0.0726062191857232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,7168,0.2492826779683431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,4096,0.06074844466315376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,6144,0.21182755629221597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,51200,1.9049608442518446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,5120,0.1736053360833062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,3584,0.05350844396485222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,3072,0.04748622245258755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,2560,0.038620445463392473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,4096,0.13640799787309435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,2048,0.0329440004295773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,3072,0.1050986647605896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,3584,0.12150488959418403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,1536,0.02495199938615163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,2560,0.0896986656718784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,1024,0.018939556346999276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,768,0.015223110715548197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,1024,0.042972445487976074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,512,0.011320888996124268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,768,0.0348364445898268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,256,0.01033155537313885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,1536,0.059233778052859835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,512,0.027833776341544256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,128,0.007970666719807519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,256,0.0229413327243593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,64,0.006991110742092133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,4096,32,0.007326222128338284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,128,0.021012443635198805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,64,0.020925333102544148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,32,0.02087466584311591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,4096,2048,0.07177511188719007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,65536,0.7601991229587131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,51200,0.581667529212104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,16384,0.20680622259775797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,12288,0.1623306671778361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,16384,0.5654364691840278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,12288,0.4271048969692654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,10240,0.12154489093356663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,8192,0.0937653316391839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,10240,0.3361955483754476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,7168,0.08503022458818223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,65536,2.2796221839057074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,51200,1.8128577338324652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,8192,0.2703111171722412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,6144,0.07178577449586657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,5120,0.061627553568945996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,7168,0.23649867375691733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,4096,0.0491902232170105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,6144,0.2017644378874037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,5120,0.16525510946909586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,3584,0.045167999135123364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,3072,0.03824533356560601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,4096,0.13000978363884821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,2560,0.03272177775700887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,3584,0.11476266384124756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,2048,0.026752000053723652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,3072,0.10060711039437188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,2560,0.08532710870107015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,1536,0.022335999541812476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,2048,0.06923111279805501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,1024,0.015840889679061044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,768,0.012254222399658628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,1536,0.055247111452950366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,512,0.01035377797153261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,1024,0.040992889139387347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,256,0.00833422193924586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,768,0.03418311145570543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,256,0.021727111604478624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,128,0.007983110845088959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,64,0.007353777686754863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,128,0.020145777199003432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3584,32,0.007304000357786815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,64,0.020041777027977836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,32,0.019722667005327012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3584,512,0.02645955483118693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,65536,0.5650586552090114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,51200,0.45643287234836155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,16384,0.16036266750759548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,16384,0.5210053126017252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,12288,0.11257333225674099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,65536,2.0860044691297746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,51200,1.6449386808607314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,10240,0.11724089251624213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,8192,0.09143555826610988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,12288,0.382022221883138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,7168,0.07950311236911349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,10240,0.31994665993584526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,8192,0.25654223230150014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,6144,0.06706666946411133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,7168,0.22513511445787218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,5120,0.0572648909356859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,4096,0.045463999112447105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,6144,0.19150222672356498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,3584,0.04110222061475118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,5120,0.156732439994812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,3072,0.03654933306905959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,2560,0.031657778554492526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,4096,0.12462933858235677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,3584,0.10986222161187066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,2048,0.02604266670015123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,3072,0.09468444188435872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,1536,0.021008888880411785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,2048,0.06478044721815321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,1024,0.014559111661381192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,2560,0.0805564456515842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,768,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,1536,0.05037777622540792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,512,0.009521777431170145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,256,0.006981333096822103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,1024,0.03802488909827338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,128,0.006588444527652528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,512,0.02514933380815718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,768,0.030962665875752766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,64,0.005919111271699269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,3072,32,0.006260444306664997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,256,0.0203137778573566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,128,0.018577777677112155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,64,0.018238221605618794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,3072,32,0.018239110708236694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,65536,0.490306642320421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,51200,0.38711733288235134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,16384,0.1378320058186849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,12288,0.10139199760225083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,16384,0.4781155586242676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,10240,0.08125333653555976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,12288,0.3643688890669081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,8192,0.06632266441980998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,10240,0.3022711012098524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,51200,1.5304817623562281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,7168,0.0609111123614841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,6144,0.051334222157796226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,8192,0.2429306772020128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,65536,1.9219733344184027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,5120,0.0437155564626058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,7168,0.2128666639328003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,4096,0.03584444522857666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,3584,0.03233866559134589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,6144,0.18154666158888075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,3072,0.02915466825167338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,4096,0.11730133162604438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,2560,0.025207999679777358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,3584,0.1040595571200053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,3072,0.0889964434835646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,2048,0.021623111433453027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,2560,0.07506844732496473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,1536,0.016421332955360413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,2048,0.061075554953681103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,1024,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,1536,0.04890844557020399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,768,0.010758221977286868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,1024,0.036024888356526695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,768,0.028901331954532202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,512,0.010055999788973067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,5120,0.14849421713087294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,256,0.008048888709810045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,128,0.005673777725961473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,512,0.023697778582572937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,64,0.005312888986534542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,256,0.019518221418062847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2560,32,0.005290666802061929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,128,0.017975111802419026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,64,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2560,32,0.017263111140992906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,65536,0.3895422352684869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,51200,0.3056551085578071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,16384,0.11090311076905991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,16384,0.4539990954928928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,12288,0.07738666401969062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,65536,1.8042560153537328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,12288,0.3414542145199246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,51200,1.417093382941352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,10240,0.0709857808219062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,8192,0.05381155676311917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,7168,0.04783644609981113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,10240,0.2866613335079617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,8192,0.23006754451327852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,6144,0.04280799958440992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,5120,0.03495733274353875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,7168,0.20217243830362955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,6144,0.17142755455440947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,5120,0.14003200001186794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,3584,0.026211554805437725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,4096,0.11043466462029351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,3072,0.02440888848569658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,3584,0.09786578019460042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,2560,0.02049066623051961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,3072,0.08423822455936009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,2048,0.019511111908488803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,2560,0.07266844643486871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,1536,0.014521777629852295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,2048,0.05915111303329468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,1024,0.011374221907721626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,1536,0.04622311062282986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,768,0.009014222357008193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,4096,0.029888000753190782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,512,0.008422222402360704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,1024,0.03293600016170078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,768,0.02795555525355869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,256,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,128,0.0052977779673205475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,64,0.005048888838953442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,512,0.02292622294690874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,128,0.016300444801648457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,2048,32,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,64,0.01590755581855774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,65536,0.3194764455159505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,51200,0.2524204519059923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,256,0.01819644371668498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,16384,0.08397066593170166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,2048,32,0.015871110889646742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,12288,0.06739377975463867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,65536,1.6617333094278972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,16384,0.42220266660054523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,12288,0.32010043991936576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,10240,0.05993244383070204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,8192,0.04683199856016371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,51200,1.3076071209377713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,7168,0.04016000032424927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,6144,0.03365600109100342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,8192,0.2164426644643148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,7168,0.1898311111662123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,10240,0.26898490058051217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,6144,0.16067200236850315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,4096,0.02409244411521488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,3584,0.02151022189193302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,4096,0.10341511170069377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,5120,0.13162222173478869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,3072,0.0192248887485928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,2560,0.01694222291310628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,3584,0.09136711226569282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,2048,0.014236445228258768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,3072,0.07981155316034953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,2560,0.06772355238596599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,1536,0.011546666423479715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,1024,0.008735111190213097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,2048,0.05472977624999153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,768,0.007646222081449296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,1536,0.042340444193945996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,512,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,1024,0.031056000126732722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,256,0.0052151112920708126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,768,0.02609599961174859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,5120,0.030307554536395605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,512,0.021353777911927965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,128,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,64,0.004259555704063839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,128,0.015204444527626038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,256,0.01717688971095615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1536,32,0.004555555681387584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,64,0.01514311134815216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,65536,0.23602843284606934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1536,32,0.015145777000321282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,51200,0.18882311715020073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,16384,0.05950044261084663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,16384,0.40010754267374676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,12288,0.04745422138108147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,51200,1.2271493275960286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,12288,0.3036293453640408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,65536,1.5661226908365886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,10240,0.038226667377683855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,8192,0.03158400124973721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,7168,0.027466666367318895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,10240,0.25445956654018825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,6144,0.024920889072948035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,8192,0.20549510584937203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,5120,0.021151999632517498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,7168,0.17923555109235975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,6144,0.15221156014336482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,4096,0.017295110556814406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,3584,0.01575999955336253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,5120,0.12448533376057942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,3072,0.014522666732470194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,4096,0.09771644406848484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,2560,0.01240711079703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,3584,0.08625688817765977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,2048,0.010480889015727572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,3072,0.07533599932988484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,2560,0.06337689028845893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,1536,0.00902133352226681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,1024,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,1536,0.03978133201599121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,2048,0.05103200011783176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,768,0.006849777781301075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,512,0.005617777920431561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,1024,0.029288887977600098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,256,0.0053093334039052325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,512,0.01996888882584042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,768,0.024879111184014216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,128,0.00453244439429707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,64,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,256,0.015879111157523263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,128,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,64,0.013841778039932251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,1024,32,0.004304000073009067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,1024,32,0.013792888985739814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,51200,0.1420382261276245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,65536,0.18954311476813424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,16384,0.04714133342107137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,65536,1.4943644205729167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,12288,0.03565866748491923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,51200,1.1709563997056749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,10240,0.02950577934583028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,8192,0.023335110810067918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,16384,0.38279022110833066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,12288,0.29063378440009224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,8192,0.1961991124682956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,10240,0.24364622433980307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,6144,0.01735555628935496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,5120,0.015646222564909194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,6144,0.1457902193069458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,7168,0.1721564398871528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,4096,0.012734221915404001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,3584,0.01165600038237042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,5120,0.11884621779123943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,3072,0.010370666782061258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,4096,0.0936773353152805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,2560,0.009369778136412302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,3584,0.08229333162307739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,2048,0.00831822223133511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,3072,0.07174133592181735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,1536,0.007335999773608313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,2560,0.06033600038952298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,7168,0.020998222960366145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,2048,0.04831199844678243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,768,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,1536,0.03765689002143012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,1024,0.027505778604083594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,512,0.004970666848950916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,768,0.023716444770495098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,256,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,128,0.003975999852021535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,512,0.019274666905403137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,64,0.0036391110883818734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,256,0.015390222271283468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,32,0.003622222277853224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,128,0.013849777479966482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,64,0.013220444321632385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,65536,0.1668764485253228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,51200,0.1326835552851359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,768,1024,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,16384,0.04490400022930569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,16384,0.3723759916093614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,12288,0.03514133228196038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,768,32,0.013314666847387949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,65536,1.4453680250379775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,51200,1.1331173578898113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,10240,0.027486221657858953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,8192,0.023136888941129048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,12288,0.2824195490943061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,7168,0.01924088928434584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,10240,0.23698578940497506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,8192,0.1910017728805542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,6144,0.01792355544037289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,5120,0.015276443627145557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,7168,0.16724356015523276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,6144,0.14169243971506754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,4096,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,3584,0.010983111129866706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,5120,0.1155360009935167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,3072,0.010791999598344168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,4096,0.09028000301784939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,2560,0.009718221922715506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,3584,0.07980355289247301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,2048,0.007663999994595845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,3072,0.06950222121344672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,2560,0.05801155832078722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,1536,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,1024,0.005973333285914526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,2048,0.046562665038638644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,768,0.004896000027656555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,1536,0.0354639987150828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,512,0.004579555657174853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,1024,0.026869333452648584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,256,0.003951111187537511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,768,0.023000889354281958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,512,0.018580444984965853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,128,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,64,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,256,0.01458666721979777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,128,0.01311555587583118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,512,32,0.003617777592606015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,64,0.012840000291665396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,65536,0.09886666801240708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,512,32,0.012472000386979846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,51200,0.07929244306352404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,16384,0.03194933467441135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,16384,0.3613457679748535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,12288,0.021651556094487507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,51200,1.100647078620063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,10240,0.01812533371978336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,12288,0.2741599877675374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,65536,1.4049386978149414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,8192,0.01956266661485036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,7168,0.01752088963985443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,10240,0.23052088419596353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,8192,0.1856515540017022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,7168,0.1623635556962755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,5120,0.012409778104888068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,6144,0.1373857789569431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,4096,0.008638222184446123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,5120,0.11243644025590684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,3584,0.008727110922336578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,4096,0.08786222007539536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,3072,0.007271111011505127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,3584,0.07715111308627658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,2560,0.006931555353932911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,3072,0.06627111302481757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,2048,0.0058853332367208265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,2560,0.055958224667443164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,6144,0.015432000160217285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,1536,0.005583999885453119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,1024,0.004879999905824661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,2048,0.04485511117511325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,768,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,1536,0.03476622369554307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,512,0.0038737778862317405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,1024,0.02611911131276025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,256,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,768,0.022275555464956496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,128,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,512,0.01791733337773217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,64,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,256,0.013787555197874704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,256,32,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,128,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,65536,0.0699146654870775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,32,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,51200,0.05794488721423679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,16384,0.022970666488011677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,16384,0.3524444368150499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,12288,0.01628533336851332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,51200,1.0721946292453342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,256,64,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,10240,0.01516888870133294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,8192,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,65536,1.3692497677273219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,7168,0.01129422254032559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,12288,0.2676355573866102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,6144,0.010072000324726105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,10240,0.2245697710249159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,5120,0.009134222235944536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,8192,0.18101599481370714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,4096,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,7168,0.15853244728512234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,6144,0.13400977187686497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,5120,0.10913066731558906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,3584,0.007680888805124495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,3072,0.006681777950790193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,4096,0.08504266871346368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,2560,0.006229333165619109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,3584,0.07443910837173462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,2048,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,3072,0.06527288754781087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,2560,0.0535786681705051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,1536,0.005030222237110138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,1024,0.004268444246715969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,2048,0.04285511043336657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,768,0.003927999900446998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,1536,0.033879111210505165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,512,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,1024,0.025459556115998164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,768,0.02129688858985901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,512,0.017885333961910672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,128,0.0029377777957253983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,256,0.01348000019788742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,64,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,128,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,32,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,64,0.01185955521133211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,65536,0.06701511144638062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,128,32,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,51200,0.05411822266048855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,128,256,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,16384,0.01774044500456916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,16384,0.35201332304212785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,51200,1.0713635550604927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,12288,0.013497778111033969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,65536,1.3690097596910265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,10240,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,8192,0.010629333555698395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,12288,0.26712356673346627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,7168,0.009756444229020013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,8192,0.18129421605004203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,6144,0.009682666924264696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,7168,0.15834666623009577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,5120,0.008643555144468943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,6144,0.13416355186038548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,5120,0.10843466387854682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,4096,0.007941332956155142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,3584,0.007032000356250339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,4096,0.08471555842293634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,3072,0.0063288890653186375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,10240,0.22441956732008192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,2560,0.005904888941182031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,3584,0.0750524467892117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,2048,0.0052666668262746595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,3072,0.06406578090455797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,1536,0.00488711107108328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,2560,0.05343466666009691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,1024,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,2048,0.042832887834972806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,1536,0.03378933336999681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,768,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,512,0.003621333175235324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,1024,0.025079111258188885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,768,0.02166844407717387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,256,0.0033440000067154565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,128,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,512,0.018002667360835604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,64,0.0029315555261241067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,256,0.013538666897349887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,64,32,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,128,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,65536,0.06491733259624906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,32,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,64,64,0.011528000235557556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,51200,0.05173244410090976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,16384,0.014392000105645923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,12288,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,16384,0.3519279956817627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,12288,0.26729954613579643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,10240,0.010393777655230628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,8192,0.009416888985368941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,10240,0.2240062289767795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,51200,1.0712683995564778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,7168,0.00868622213602066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,6144,0.008316444853941599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,8192,0.1814568837483724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,65536,1.3687271542019313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,5120,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,4096,0.009362666971153682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,7168,0.15817422337002224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,3584,0.008644444247086843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,6144,0.13405956162346735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,5120,0.10872355434629653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,3072,0.008336889247099558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,2560,0.0069724445541699724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,4096,0.08495022190941705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,3584,0.07480800151824951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,2048,0.005495999836259418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,3072,0.06436622142791748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,1536,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,2560,0.053898665640089244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,1024,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,2048,0.043098668257395424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,768,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,1536,0.033032887511783175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,512,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,1024,0.025401777691311304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,256,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,768,0.02162844439347585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,128,0.0028213332924577924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,512,0.017643555998802185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,256,0.013561777770519257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,128,0.012072888513406118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3073,32,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,64,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3073,32,32,0.011415111521879831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,16384,3.150499555799696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,51200,9.635021633572048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,16384,4.388970692952474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,12288,2.3666932847764754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,16384,3.249086168077257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,51200,15.074118720160591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,51200,10.386448330349392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,10240,1.9563013712565105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,12288,3.0318781534830728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,10240,2.479729758368598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,12288,2.4148409101698136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,8192,1.5728800031873915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,10240,2.0113324059380426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,8192,1.918966293334961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,7168,1.337259504530165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,8192,1.6134533352322047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,6144,1.1487777498033311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,7168,1.6747271219889324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,7168,1.4377920362684462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,6144,1.4186311297946508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,5120,0.9164898130628797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,4096,0.7434408929612902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,5120,1.1794355180528429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,6144,1.226793819003635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,5120,1.029334174262153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,3584,0.6942470868428549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,4096,0.9285484949747721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,4096,0.8356942070855035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,3584,0.8169537650214301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,3072,0.5556649102105035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,3072,0.7050559785630969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,3584,0.7427262200249566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,3072,0.6499173376295302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,2048,0.4072631200154622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,2560,0.600088013543023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,2560,0.5548933347066244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,2560,0.4651341968112522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,1536,0.306679990556505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,2048,0.49333424038357204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,2048,0.46245066324869794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,1024,0.19060533576541475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,1536,0.3880106608072917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,1536,0.37014399634467227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,1024,0.2935875521765815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,768,0.15238132741716173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,1024,0.28329600228203666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,512,0.11958666642506917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,768,0.2329742113749186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,256,0.09061066971884833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,512,0.1891768905851576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,512,0.19922222031487358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,128,0.07755377557542589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,256,0.17652177810668945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,256,0.14813955624898276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,64,0.0671715537707011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,128,0.16801955964830187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,65536,128,0.11618933412763809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,65536,32,0.06951377789179485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,64,0.1694942182964749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,32,0.17331466409895155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,65536,768,0.23217333687676323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,51200,7.513777838812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,65536,9.67215813530816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,65536,15.381627400716146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,65536,10.741675482855904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,16384,2.4129341973198786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,16384,3.5211289723714194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,51200,11.983019510904947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,12288,1.8814951578776042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,16384,2.629596498277452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,51200,8.368646409776476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,12288,2.4157297346327042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,10240,1.4992248747083876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,10240,1.9796070522732203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,8192,1.1620719697740343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,12288,1.9785387251112196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,10240,1.647762616475423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,7168,1.0763457616170247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,8192,1.5373591317070856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,7168,1.3416106965806749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,8192,1.3176320393880208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,6144,0.9422079722086588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,7168,1.1612106959025066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,5120,0.8041840129428439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,6144,1.0009021759033203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,5120,0.9422818289862739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,6144,1.1322960323757596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,4096,0.5775858031378852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,5120,0.8419626553853353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,4096,0.7422195540534126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,3584,0.5075404379102919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,4096,0.6839271121554904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,3584,0.6517742474873861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,3072,0.43639998965793186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,3584,0.6070319811503092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,3072,0.5638471179538304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,2560,0.37485422028435605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,3072,0.5293342272440592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,2560,0.48035287857055664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,2048,0.2949777709113227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,2560,0.45378579033745664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,2048,0.3939777745140924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,1536,0.24462310473124185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,2048,0.3775057792663574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,1024,0.15526489416758218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,1536,0.3113822142283122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,1536,0.3024560080634223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,768,0.1307564444012112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,1024,0.23581332630581328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,1024,0.23142488797505698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,512,0.09540977742936875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,512,0.15936711099412706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,512,0.15564444330003527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,256,0.07261955738067627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,768,0.19168621963924834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,128,0.06087733639611138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,768,0.18732710679372153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,256,0.12181599934895833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,256,0.14123555024464926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,64,0.05189955565664503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,51200,32,0.05270666546291775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,128,0.13383289178212485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,64,0.1357004510031806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,51200,32,0.1376711130142212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,51200,128,0.09681333435906304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,51200,2.619906743367513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,65536,3.1537759568956165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,65536,5.712325202094184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,65536,4.6814248826768665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,16384,0.7822533183627658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,51200,4.531586541069879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,16384,1.2959262000189888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,51200,3.6558587816026478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,12288,0.5988906754387749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,16384,1.182779524061415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,10240,0.4882328775193956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,12288,0.9289431042141385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,10240,0.7669795354207357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,12288,0.8929689195421008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,10240,0.7492382261488172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,8192,0.6059768994649252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,7168,0.33926932017008465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,8192,0.6069555812411839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,7168,0.5317191017998589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,6144,0.30464445220099556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,8192,0.4235057830810547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,5120,0.2579591009351942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,7168,0.532535129123264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,6144,0.4527200063069661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,4096,0.19605333275265166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,6144,0.46018399132622617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,5120,0.37252622180514866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,5120,0.38621245490180117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,3584,0.18053332964579263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,4096,0.295106675889757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,4096,0.3121359878116184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,3072,0.1490106715096368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,3584,0.27497511439853245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,3584,0.25946222411261666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,2560,0.1324817736943563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,3072,0.22522134251064727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,3072,0.24275554551018608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,2048,0.10190666384167141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,2560,0.20791644520229766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,2048,0.15923466947343615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,1536,0.08004799816343519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,2048,0.1743333339691162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,1024,0.0577404432826572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,2560,0.19374222225613066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,1536,0.12620888815985784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,1536,0.14103910658094618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,768,0.04428977767626444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,1024,0.0957368877198961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,768,0.07646666632758246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,768,0.09070666631062825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,1024,0.10938044389088948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,512,0.03985866573121812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,256,0.03157511022355821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,512,0.07551733652750652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,512,0.06492177645365398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,256,0.055762668450673424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,128,0.022335999541812476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,64,0.01923199991385142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,128,0.04774666825930277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,128,0.05227288934919569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,16384,256,0.05955288807551066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,16384,32,0.019644444187482197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,64,0.053120000494851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,16384,32,0.05418489045566983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,65536,2.35098754035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,51200,1.8309120602077906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,65536,4.592671288384332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,65536,4.004869249131945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,16384,0.6034871207343208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,16384,1.0473475986056857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,12288,0.4443511168162028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,51200,3.6575838724772134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,12288,0.757712894015842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,51200,3.1388763851589627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,10240,0.3783857822418213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,16384,1.0210338168674045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,8192,0.3099457687801785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,10240,0.6259804301791722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,10240,0.6462533209058974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,7168,0.26181332270304364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,12288,0.7688088946872287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,8192,0.4981706407335069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,8192,0.5225537618001302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,6144,0.22636710272894967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,7168,0.4366168975830078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,6144,0.3716533448961046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,7168,0.46124177508884007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,5120,0.18411289321051705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,4096,0.146525330013699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,6144,0.39674311214023167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,5120,0.30658666292826336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,5120,0.3306728998819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,3584,0.13184000386132136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,4096,0.24139822853936088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,4096,0.26928888426886666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,3584,0.2122346692615085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,3584,0.23744887775845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,2560,0.0969724456469218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,3072,0.20881599850124785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,3072,0.18545067310333252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,2560,0.1587351163228353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,2048,0.07549244165420532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,2560,0.17941155698564318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,1536,0.06510044468773736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,2048,0.1314142280154758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,2048,0.15173867013719347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,1536,0.10427822007073297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,1536,0.12210933367411296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,1024,0.044421331750022046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,768,0.0347635547320048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,1024,0.07883466614617242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,1024,0.09440977705849542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,512,0.024989333417680528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,768,0.07899200254016452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,512,0.054004444016350635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,256,0.019761777586407132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,768,0.06366044282913208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,512,0.06577244732115003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,128,0.017880888448821176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,256,0.045610666275024414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,256,0.051776889297697276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,128,0.04168355464935303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,64,0.015621332658661736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,12288,128,0.0417902237839169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,32,0.015283554792404175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,12288,3072,0.11149156093597412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,32,0.041631112496058144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,12288,64,0.04180711176660326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,65536,2.2115760379367404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,51200,1.7501537534925673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,65536,4.120079040527344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,16384,0.541719118754069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,65536,3.6162469651963978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,51200,3.2027467091878257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,16384,0.9338959587944878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,12288,0.43749332427978516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,16384,0.9218862321641711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,10240,0.33780799971686465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,12288,0.6706879933675131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,51200,2.8386027018229165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,12288,0.6944693459404839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,10240,0.5547351307339138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,8192,0.30172710948520237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,10240,0.5867644415961372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,7168,0.23179290029737684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,8192,0.44320178031921387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,8192,0.47451824612087673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,7168,0.3883439964718289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,6144,0.18973067071702746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,7168,0.4269244405958388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,5120,0.1566186613506741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,6144,0.3315315511491564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,4096,0.12578133742014566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,5120,0.2718755669063992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,6144,0.3603208859761556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,3584,0.11847021844651963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,5120,0.3003928926255968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,4096,0.21595377392239043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,4096,0.2434177663591173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,3072,0.09567111068301731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,3584,0.1896453301111857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,3584,0.21567822827233207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,2560,0.08499022324879964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,3072,0.1651795572704739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,3072,0.190010666847229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,2048,0.06947111421161227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,2560,0.16224711471133763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,2560,0.1424702273474799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,2048,0.11751466327243382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,1536,0.05190666516621908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,2048,0.13665689362419978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,1024,0.03839110996988084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,1536,0.09317511320114136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,1536,0.1119964387681749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,768,0.029161777761247423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,1024,0.07099999984105428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,1024,0.08619466755125259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,512,0.025032000409232244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,768,0.07221421930525038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,768,0.056913779841529004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,512,0.04733511143260532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,256,0.01851111153761546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,512,0.06084089146720039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,128,0.015471999843915304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,256,0.03899733225504557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,256,0.04782488942146301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,128,0.03516355488035414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,64,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,10240,128,0.03880000114440918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,10240,32,0.016192000773217943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,64,0.03478666808870103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,10240,32,0.034862223598692156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,65536,1.712747573852539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,51200,1.2704586452907987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,65536,3.4909324645996094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,16384,0.4182071155971951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,65536,3.358222325642904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,16384,0.8168720139397515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,12288,0.3179617722829183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,16384,0.8569537798563639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,51200,2.774280971950955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,12288,0.5964915487501357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,10240,0.26710489061143666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,12288,0.6482257843017578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,8192,0.21817955705854628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,10240,0.48725154664781356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,51200,2.6253689659966364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,10240,0.5442986488342285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,8192,0.38813243971930605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,7168,0.1861564450793796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,8192,0.44131022029452854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,6144,0.15818044874403211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,7168,0.38873688379923504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,7168,0.34171377287970645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,6144,0.2904728783501519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,5120,0.13964888784620497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,6144,0.3329431215922038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,4096,0.11253066857655843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,5120,0.27647466129726833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,5120,0.23967822392781576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,4096,0.18825956185658774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,4096,0.2247075504726834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,3072,0.0791084435251024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,3584,0.19855911201900908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,3584,0.16727733612060547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,3072,0.145541336801317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,2560,0.06443822383880615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,3072,0.1748960018157959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,2048,0.05322666631804573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,2560,0.12485333283742268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,2560,0.1509768830405341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,2048,0.10367466343773736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,1536,0.042706665065553456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,2048,0.12696977456410727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,3584,0.09063911437988281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,1024,0.0314408904976315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,1536,0.08231200112236871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,768,0.023997333314683702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,1536,0.10371466477711995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,1024,0.06193511353598701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,1024,0.07973600096172757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,512,0.018497778309716117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,768,0.04898666673236423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,256,0.014538667268223233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,768,0.06730577680799696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,512,0.056272890832689076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,512,0.039727999104393855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,128,0.016454221473799813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,256,0.032405333386527166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,256,0.04433066646258036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,64,0.014114666316244336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,128,0.029834667841593426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,8192,128,0.035783999496036105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,8192,32,0.014832889040311178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,32,0.029193778832753498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,8192,64,0.030059556166330974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,65536,1.4891671074761286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,51200,1.190460417005751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,65536,3.2022516462537975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,65536,3.0988025665283203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,51200,2.563311047024197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,16384,0.35767287678188747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,51200,2.4240169525146484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,16384,0.7416622373792859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,12288,0.2649306721157498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,16384,0.7951955795288086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,12288,0.5438790851169163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,10240,0.23953421910603842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,12288,0.6001555654737685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,8192,0.19150222672356498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,10240,0.4516311221652561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,10240,0.5082391103108724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,7168,0.1639431052737766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,8192,0.36871200137668186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,8192,0.40905332565307617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,6144,0.14065244462754992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,7168,0.317392004860772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,7168,0.3607031239403619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,6144,0.2702782154083252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,5120,0.11852533287472194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,6144,0.3100026713477241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,5120,0.221343994140625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,4096,0.0967786643240187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,5120,0.25716622670491535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,3584,0.08314133352703519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,4096,0.1740800009833442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,4096,0.20823821756574842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,3072,0.07267822159661187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,3584,0.18420799573262533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,3584,0.15451289547814265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,3072,0.1357626650068495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,2560,0.0595395565032959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,3072,0.1622426642311944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,2048,0.05036711030536228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,2560,0.11618578433990479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,2560,0.1406195561091105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,1536,0.037647111548317805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,2048,0.09624177879757351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,2048,0.11843910482194688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,1024,0.026191110412279766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,1536,0.09614311324225532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,1536,0.07712977462344699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,1024,0.05786755349900988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,768,0.020291555258962844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,1024,0.0743760002983941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,512,0.016224000189039443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,768,0.045231110519833036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,256,0.012934221989578672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,512,0.03645599881807963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,768,0.06294311417473687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,512,0.053526222705841064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,128,0.013527111046844058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,256,0.04007733199331496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,256,0.029662221670150757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,64,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,7168,32,0.011712888876597086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,7168,128,0.03330488999684652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,128,0.029213332467608984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,64,0.027477333943049114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,7168,32,0.027144889036814373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,51200,0.951816029018826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,65536,1.161255094740126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,65536,2.9332035912407766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,16384,0.29779288503858775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,51200,2.338075637817383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,65536,2.9656319088406033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,12288,0.22402400440639922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,51200,2.3183199564615884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,16384,0.683718204498291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,16384,0.7590782377454969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,10240,0.1898097727033827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,12288,0.5023031234741211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,8192,0.15410133202870688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,12288,0.5745644569396973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,10240,0.4174284405178494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,8192,0.3344382180107965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,7168,0.13340977827707926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,10240,0.4838542408413357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,8192,0.3914968967437744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,6144,0.11707377433776855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,7168,0.2937964333428277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,7168,0.34528801176283097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,6144,0.24975999196370444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,5120,0.09468533595403035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,6144,0.2950231234232585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,4096,0.07582488987180921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,5120,0.20475822024875215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,5120,0.24598577287462023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,3584,0.06857688559426202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,4096,0.1987146668963962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,3072,0.060059554047054715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,3584,0.14205511411031088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,3584,0.176545778910319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,2560,0.05071555574735006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,3072,0.12452444765302871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,3072,0.15539111031426325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,2048,0.0406675570540958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,2560,0.10781866974300808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,2560,0.13379022810194227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,2048,0.08923555745018853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,1536,0.030652443567911785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,1536,0.07065244515736897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,4096,0.16236800617641872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,1024,0.023157333334287006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,1536,0.09228622251086765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,1024,0.07150044706132677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,1024,0.052481777138180204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,768,0.017986666825082567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,2048,0.1126204464170668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,768,0.06040800280041165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,512,0.01404622197151184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,768,0.04244800077544319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,256,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,512,0.033794667985704206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,256,0.03777066535419888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,256,0.027552000350422327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,128,0.010055999788973067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,128,0.031369778845045306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,64,0.009039999710188972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,128,0.025812443759706285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,6144,32,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,64,0.025432000557581585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,6144,32,0.025116443634033203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,6144,512,0.05042666527960035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,65536,1.0836586422390408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,51200,0.8174035814073352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,65536,2.668592029147678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,65536,2.8340382046169705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,16384,0.27219290203518337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,51200,2.0988356272379556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,12288,0.20988799466027153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,16384,0.6208835707770454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,16384,0.726187547047933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,10240,0.1779902246263292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,12288,0.46048710081312394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,12288,0.5495244661966959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,8192,0.14618933200836182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,10240,0.38347734345330137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,10240,0.4621991051567926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,51200,2.2173794640435114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,7168,0.11946045027838813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,8192,0.30733956231011283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,6144,0.10659199953079224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,8192,0.3752151065402561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,7168,0.3311519887712267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,6144,0.23014044761657715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,7168,0.27082843250698513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,6144,0.28295644124348956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,4096,0.06631288925806682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,5120,0.18813955783843994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,4096,0.14794845051235622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,5120,0.23526488410101998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,3584,0.05703644620047676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,4096,0.18919732835557726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,3072,0.05212000012397766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,3584,0.16907733016543916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,3584,0.13170488675435385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,3072,0.1142151090833876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,2560,0.04270044300291273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,3072,0.1480613284640842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,5120,0.07991733153661092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,2048,0.03755199909210205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,2560,0.09739200274149577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,1536,0.027638221780459087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,2560,0.1283751063876682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,2048,0.10845333337783813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,1024,0.020153777466879953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,1536,0.06494577725728352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,1536,0.08852533499399821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,1024,0.04772888951831394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,1024,0.0678773323694865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,768,0.017650667164060805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,512,0.014813333749771118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,768,0.05748977926042345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,768,0.03788977861404419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,256,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,512,0.04658133453792996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,512,0.030976000759336684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,256,0.024708444873491924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,128,0.00907555553648207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,256,0.03587733374701606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,64,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,128,0.023037332627508376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,5120,32,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,2048,0.08201599783367582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,5120,128,0.03128800127241347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,32,0.022650667362742957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,65536,0.9012329313490125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,5120,64,0.02271555529700385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,51200,0.6876684294806586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,65536,2.397073745727539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,16384,0.22612267070346406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,51200,1.8966994815402562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,65536,2.7007289462619357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,12288,0.17731466558244494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,16384,0.5636755625406901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,12288,0.4260755644904242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,51200,2.113023122151693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,16384,0.6924737824334039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,10240,0.13860800531175402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,12288,0.5262071291605631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,8192,0.1127600007587009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,10240,0.3506657812330458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,10240,0.4420924451616075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,8192,0.281821330388387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,7168,0.09884533617231582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,8192,0.35743554433186847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,6144,0.08416266573799981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,7168,0.31620976659986705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,7168,0.24707643191019693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,6144,0.20926666259765625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,5120,0.0658044417699178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,6144,0.2697750992245144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,5120,0.17095822758144805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,4096,0.05725422170427111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,5120,0.22431998782687715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,3584,0.04786844385994805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,4096,0.1358835564719306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,4096,0.18108888467152914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,3072,0.042062222957611084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,3584,0.12008533212873672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,3584,0.16046755843692356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,2560,0.035338666703965925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,3072,0.10400533013873631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,3072,0.14180621835920545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,2560,0.0892088876830207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,2560,0.12168178293440078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,2048,0.029509332444932725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,1536,0.022727999422285292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,2048,0.07131377855936687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,2048,0.10219644175635444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,1024,0.017172444197866652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,1536,0.058768888314565025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,1536,0.08384178082148235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,768,0.013612444202105204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,1024,0.043135109874937266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,1024,0.06480977932612102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,768,0.034994665119383074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,768,0.05319022138913473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,512,0.012504888905419244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,256,0.010141332944234213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,512,0.027627555860413447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,512,0.043988443083233304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,128,0.010041777458455827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,256,0.022896000080638464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,256,0.034620443979899086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,64,0.009335999687512716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,128,0.021252445048756067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,4096,32,0.009386666946940953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,4096,128,0.02924088968171014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,64,0.020663999848895602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,4096,32,0.020647999313142564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,65536,0.7447982364230685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,51200,0.6378329065110948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,65536,2.2328008015950522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,16384,0.20083822144402397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,65536,2.569929758707682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,16384,0.533931573232015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,51200,1.7685537338256836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,12288,0.14311199718051487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,12288,0.4005902343326145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,16384,0.6614213519626194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,51200,2.0121332804361978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,12288,0.5011564360724555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,10240,0.11240355173746745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,8192,0.09320622020297581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,10240,0.33447199397616917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,10240,0.4214924441443549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,7168,0.0820924441019694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,8192,0.3416106700897217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,8192,0.26893777317470974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,6144,0.06934666633605957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,7168,0.2351466549767388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,7168,0.30175643497043186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,5120,0.05626844697528415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,6144,0.25701954629686147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,6144,0.20020088884565565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,4096,0.047437333398395114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,5120,0.1630444394217597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,5120,0.21347378359900582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,3584,0.04018933243221707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,4096,0.1724053356382582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,4096,0.12985332806905112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,3072,0.034527109728919134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,3584,0.11362844043307835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,3584,0.15354222721523708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,2560,0.029727998707029555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,3072,0.09990133179558648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,3072,0.13529600037468806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,2048,0.02444533341460758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,2560,0.11583022276560466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,2560,0.08443555566999648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,1536,0.020601777566803824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,2048,0.06773066520690918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,1024,0.014204444156752693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,2048,0.09861244757970174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,1536,0.05467289023929172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,1536,0.079949332608117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,768,0.011667555405033959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,1024,0.04052444299062093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,1024,0.06302755408816867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,768,0.03259911139806112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,512,0.009407999614874521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,768,0.05270933442645603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,256,0.007635555333561367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,512,0.02611733310752445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,512,0.04391910963588291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,256,0.021640888518757288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,128,0.007359999749395583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,256,0.03318666749530368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,64,0.006599999964237213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,128,0.019934222102165222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3584,128,0.02823022339079115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3584,32,0.006653333289755716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,64,0.019621334142155118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3584,32,0.01980266637272305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,65536,0.572407086690267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,51200,0.47297244601779515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,65536,2.161639107598199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,51200,1.6726178063286676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,16384,0.15792532761891684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,65536,2.4398409525553384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,16384,0.5052177641126845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,12288,0.10967377821604411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,16384,0.6290853288438585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,12288,0.37963554594251847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,51200,1.911267598470052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,10240,0.0934106641345554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,12288,0.4777057965596517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,8192,0.08578666713502671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,10240,0.3167697853512234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,10240,0.4008382161458333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,7168,0.06967377662658691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,8192,0.2552986674838596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,6144,0.05960977739757962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,8192,0.3258559968736437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,7168,0.22262843449910483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,5120,0.04787022206518385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,7168,0.2881191041734484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,6144,0.19027554988861084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,6144,0.24482311142815483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,4096,0.04150844282574124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,5120,0.15534133381313747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,3584,0.035106665558285184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,4096,0.12249333328670925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,5120,0.20358755853441027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,3072,0.03264088763131036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,4096,0.16433689329359266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,3584,0.10837688710954453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,3584,0.14584088325500488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,2560,0.027245332797368366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,3072,0.0944346653090583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,3072,0.12880444526672363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,2048,0.022095110681321885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,2560,0.08000266551971436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,1536,0.017864889568752713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,2048,0.09348266654544407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,2048,0.0653546651204427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,1536,0.05033422178692288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,1024,0.013815999858909182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,1536,0.07640622059504192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,768,0.011135111252466837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,1024,0.038351111941867404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,1024,0.06016177601284451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,512,0.008791111409664154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,768,0.03087200058831109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,768,0.048455999957190625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,256,0.006974221931563483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,512,0.04045333464940389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,512,0.025601777765485976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,256,0.020267556111017864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,128,0.006588444527652528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,256,0.032398223876953125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,2560,0.11120622687869602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,64,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,128,0.018584888842370775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,3072,32,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,3072,128,0.028648889727062646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,64,0.018606222338146634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,3072,32,0.018555555078718398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,65536,0.5316781997680664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,51200,0.40825332535637754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,65536,1.9211209615071614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,51200,1.5134471257527669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,16384,0.13706755638122559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,16384,0.47632445229424375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,51200,1.909683651394314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,65536,2.437772538926866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,12288,0.10185866885715061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,10240,0.09312711159388225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,16384,0.6293830871582031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,12288,0.35906222131517196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,8192,0.06618399752510919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,10240,0.3010728889041477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,12288,0.47674401601155597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,7168,0.06038844585418701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,10240,0.40184444851345485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,8192,0.24203999837239584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,6144,0.05097422334882948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,7168,0.21182400650448271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,7168,0.28707199626498753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,6144,0.17995911174350313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,5120,0.04167822334501478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,6144,0.24400001102023652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,5120,0.14699822001987034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,4096,0.03372977839575873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,5120,0.20264266596900093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,8192,0.32530400488111705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,3584,0.03199377655982971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,4096,0.16404622130923802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,3584,0.10223999950620864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,3072,0.026161778304311965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,3584,0.14606666564941406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,3072,0.08868088987138535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,2560,0.023021333747439917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,3072,0.12838844458262125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,2560,0.07530399825837877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,4096,0.1166257725821601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,2048,0.018904889623324077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,2560,0.1114399962955051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,1536,0.015192889504962497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,2048,0.09265866544511582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,1536,0.0762044456270006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,1536,0.04897866646448771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,1024,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,1024,0.058876441584693066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,1024,0.036064889695909284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,768,0.01033333357837465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,768,0.04815999997986687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,512,0.009030222064918941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,768,0.029337777031792536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,2048,0.06030133035447863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,512,0.02357599967055851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,256,0.007003555695215861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,512,0.0411208901140425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,256,0.01905866629547543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,128,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,256,0.03128977616628011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,64,0.007723555796676212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,128,0.01757866640885671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2560,128,0.026800000005298193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2560,32,0.007683555285135905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,64,0.017254221770498488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2560,32,0.017287999391555786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,65536,0.42403467496236164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,51200,0.3418426778581407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,65536,1.799207051595052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,16384,0.10926755269368489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,51200,1.407400025261773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,16384,0.44838847054375547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,12288,0.0865066647529602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,65536,2.3090489705403647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,16384,0.5956888728671604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,10240,0.07141066922081842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,12288,0.33922044436136883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,51200,1.8069377475314672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,12288,0.45367908477783203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,10240,0.28521866268581814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,10240,0.38068265385097927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,7168,0.049123555421829224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,8192,0.23016977310180664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,8192,0.30936532550387913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,6144,0.04448533389303419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,7168,0.2009404500325521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,7168,0.2724551094902886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,6144,0.1708106729719374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,5120,0.036143110858069524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,5120,0.1396906640794542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,8192,0.05986666679382324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,4096,0.029804444975323145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,5120,0.19213510884179008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,3584,0.026502221822738647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,4096,0.11021155781216091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,4096,0.15512177679273817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,3072,0.022350221872329712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,3584,0.0976782242457072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,3584,0.13780532942877874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,2560,0.019280888968043856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,3072,0.08409777614805435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,3072,0.12201600604587132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,2560,0.07190755340788099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,2560,0.1050835582945082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,2048,0.01650044487582313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,1536,0.014150222142537435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,2048,0.05901777744293213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,2048,0.08940533134672378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,1024,0.010363555616802638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,1536,0.04621511035495334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,1536,0.07233866718080309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,768,0.008718222379684448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,1024,0.033025778002209134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,1024,0.05466755562358432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,768,0.027808000644048054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,512,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,768,0.045906666252348155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,256,0.006584888945023219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,512,0.02201066745652093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,512,0.0388008885913425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,256,0.017954667409261067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,128,0.006005333529578314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,256,0.03055822186999851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,6144,0.23229689068264434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,64,0.006656000183688269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,2048,128,0.02609866691960229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,128,0.0166275550921758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,2048,32,0.006956444846259222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,64,0.016255999604860943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,2048,32,0.01628977722591824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,65536,0.3162204424540202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,51200,0.24576176537407768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,65536,1.703521728515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,51200,1.334501372443305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,16384,0.087391111585829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,65536,2.19561767578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,16384,0.4322764343685574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,12288,0.06052177482181125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,12288,0.319942209455702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,16384,0.5647199948628744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,51200,1.7066639794243708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,10240,0.05130222108629015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,12288,0.42864176962110734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,10240,0.2683520052168104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,8192,0.044215109613206655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,10240,0.36148977279663086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,7168,0.037385778294669256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,8192,0.21544978353712294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,8192,0.29257689581977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,6144,0.03243733445803324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,7168,0.18962311744689941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,7168,0.2583386633131239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,5120,0.027278222971492346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,6144,0.1607875559065077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,6144,0.21933866871727836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,4096,0.02281422250800663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,5120,0.13049599859449598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,5120,0.18148088455200195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,3584,0.021145777569876775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,4096,0.14631199836730957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,4096,0.10421866840786403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,3584,0.09033688571718003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,3072,0.018637332651350234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,3584,0.12993066840701634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,2560,0.01569422251648373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,3072,0.07940177785025702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,3072,0.11538488335079616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,2048,0.013571555415789286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,2560,0.06762578090031941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,2560,0.09946488671832615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,1536,0.010789333118332757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,2048,0.05468177795410156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,2048,0.08428800106048584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,1536,0.04275911053021749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,1536,0.06877155436409844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,768,0.007619555625650618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,1024,0.05135644475618998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,1024,0.030939555830425684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,512,0.006667555620272954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,768,0.04351466563012865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,512,0.021369778447681006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,256,0.0052915554907586836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,256,0.016882666283183627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,256,0.028903110159768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,1024,0.00871733327706655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,128,0.004639111045334074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,128,0.024710221423043147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,64,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,128,0.015615999698638916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1536,32,0.004411555412742826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,64,0.01500888831085629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1536,512,0.036892443895339966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,32,0.015169777803950839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,65536,0.21105421913994682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,51200,0.17484711276160347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1536,768,0.025782222549120586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,65536,1.5628595352172852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,51200,1.2415413326687283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,16384,0.0633902218606737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,16384,0.3990791108873155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,12288,0.04230755567550659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,16384,0.5634329054090712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,65536,2.1795804765489364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,10240,0.03676088982158237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,12288,0.3028959963056776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,51200,1.704321755303277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,8192,0.02918400035964118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,12288,0.4286373456319173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,10240,0.25501155853271484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,7168,0.026309332913822595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,10240,0.3614071210225423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,8192,0.20449154906802705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,8192,0.2925671206580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,6144,0.02273422148492601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,7168,0.17932621637980142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,7168,0.2584551175435384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,5120,0.020124443703227572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,6144,0.15167821778191462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,6144,0.22023644712236193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,4096,0.016689777374267578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,5120,0.12356177965799968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,5120,0.18242488967047799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,3584,0.015124445160230001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,4096,0.14622488286760119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,4096,0.09780799680285984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,3072,0.013749333719412485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,3584,0.13017600112491184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,3584,0.0865937802526686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,3072,0.07527022229300605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,2560,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,3072,0.11459822124905056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,2048,0.01020088874631458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,2560,0.06379733482996623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,2560,0.09999377859963311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,2048,0.0507279998726315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,1536,0.008741333252853816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,2048,0.08329333199395074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,1536,0.03887822230656942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,1024,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,1024,0.02887644370396932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,768,0.006337777607970768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,1024,0.05104000038570828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,768,0.02403733299838172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,512,0.005582222094138463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,768,0.04368266794416639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,512,0.019634667370054457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,256,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,512,0.03676889008945889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,256,0.01555377741654714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,1536,0.06863733132680257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,128,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,256,0.029135111305448744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,64,0.004295111116435793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,1024,128,0.024624889095624287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,1024,32,0.004053333153327306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,128,0.014523555835088095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,64,0.014208000567224292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,1024,32,0.013750221994188098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,65536,0.17385155624813506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,51200,0.13686045010884604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,51200,1.1705332862006295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,65536,1.4905938042534723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,16384,0.04518489042917887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,65536,2.048100365532769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,16384,0.3824924363030328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,12288,0.03563022282388475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,16384,0.5304595629374186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,12288,0.2900746663411458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,10240,0.029612445169025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,12288,0.4043004512786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,51200,1.6065858205159504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,8192,0.023642667465739783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,10240,0.24359377225240073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,10240,0.33980266253153485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,7168,0.020063999626371596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,7168,0.17187289396921793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,8192,0.19674311743842232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,8192,0.276071998808119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,6144,0.018058665924602084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,5120,0.015478221906556023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,7168,0.2433875666724311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,6144,0.14599288834465876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,5120,0.1182755496766832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,6144,0.20698400338490805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,4096,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,3584,0.011790222591824003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,5120,0.1706542174021403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,4096,0.09342666467030843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,3072,0.01070844464831882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,4096,0.138319108221266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,3584,0.08246755599975586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,2560,0.00959022260374493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,3584,0.12310311529371475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,3072,0.07121333148744372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,3072,0.10797066820992364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,2048,0.008329778081840938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,2560,0.06072355641259087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,2560,0.09352266788482666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,1536,0.007296000089910295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,2048,0.07781955268647936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,2048,0.0480640000767178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,1024,0.006014222072230444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,1536,0.06481955448786418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,1024,0.027698665857315063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,1024,0.04798044429885017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,768,0.005294222384691238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,512,0.004888888862397936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,768,0.041144887606302895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,768,0.023032888770103455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,256,0.003984000119898054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,512,0.019311999281247456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,512,0.03465955456097921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,128,0.003899555653333664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,256,0.014879999889267815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,256,0.028151111470328435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,64,0.0035946667194366455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,128,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,768,128,0.023393778337372675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,768,32,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,65536,0.14515111181471083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,32,0.013119999733236102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,64,0.013653332988421122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,768,1536,0.036751998795403376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,51200,0.11322311560312907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,65536,1.4438915252685547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,51200,1.1324035856458876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,16384,0.03980266716745164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,65536,2.0475502014160156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,16384,0.3718675507439508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,12288,0.029947555727428857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,16384,0.5306417677137587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,12288,0.2824186748928494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,10240,0.023805333508385554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,51200,1.6009484397040472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,12288,0.40361865361531574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,8192,0.019173334042231243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,10240,0.2371759944491916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,10240,0.3390764395395915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,7168,0.01737244427204132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,8192,0.19140533606211343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,8192,0.2745280000898573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,7168,0.16714844438764784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,6144,0.014987554815080432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,7168,0.2424088848961724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,6144,0.14155822330051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,5120,0.01331022216214074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,6144,0.2067377832200792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,4096,0.011009777585665384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,5120,0.11517333984375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,5120,0.170470224486457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,3584,0.010075555907355415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,4096,0.09042133225335015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,3584,0.08010488748550415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,3072,0.009400000174840292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,3584,0.12221599949730767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,3072,0.06955466667811076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,3072,0.10733422305848862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,2560,0.008403555386596257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,2048,0.00777422222826216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,2560,0.09294844335979885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,2560,0.05843466520309448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,1536,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,4096,0.13701600498623318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,2048,0.04608622193336487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,1024,0.00565155554148886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,2048,0.07786755429373847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,1536,0.03678044345643785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,1536,0.06256622076034546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,768,0.004976888911591636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,1024,0.026787555880016748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,1024,0.04840177628729078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,512,0.004622222234805425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,768,0.022321777211295232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,256,0.004307555655638377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,768,0.04089511103100247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,512,0.03444800111982558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,256,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,128,0.004022222426202563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,256,0.027425777581002977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,64,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,512,128,0.023042667243215773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,512,32,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,64,0.012828444441159567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,32,0.01257777793539895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,65536,0.10667555862002903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,512,0.018982221682866413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,51200,0.08548444509506226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,512,128,0.013097777962684631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,65536,1.4046292834811742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,16384,0.0347671111424764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,16384,0.361241790983412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,51200,1.1004373762342665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,12288,0.02532088922129737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,65536,2.0333599514431424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,51200,1.5924480226304796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,16384,0.5268408987257216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,12288,0.27449509832594127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,12288,0.40026577313741046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,8192,0.018613333503405254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,10240,0.23011999660068086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,10240,0.33723288112216526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,8192,0.18542222181955972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,8192,0.27330756187438965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,7168,0.16239110628763834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,6144,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,7168,0.2414933310614692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,10240,0.020580444071027968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,5120,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,6144,0.20510133107503256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,6144,0.137335115008884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,5120,0.11198666360643174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,4096,0.008207110895050896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,5120,0.16935022672017416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,4096,0.0873982244067722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,7168,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,3584,0.007694222033023834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,3072,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,4096,0.13602399826049805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,3584,0.07650399870342679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,3584,0.12159289254082574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,3072,0.06642488638559978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,3072,0.10674399799770778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,2560,0.05653866794374254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,2048,0.005647999958859549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,2560,0.091867552863227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,2048,0.044638223118252225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,2048,0.07706666655010648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,1536,0.005264889034960005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,1536,0.0622239973809984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,2560,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,1536,0.03472977876663208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,1024,0.004643555730581284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,768,0.00425155543618732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,1024,0.04771733283996582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,1024,0.026103999879625108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,768,0.02205955485502879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,512,0.0036888888312710654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,768,0.04041422075695462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,256,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,512,0.018248000078731112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,512,0.034078223837746516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,128,0.003608889049953885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,256,0.027494221925735474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,256,0.014154665999942355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,64,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,128,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,256,32,0.0032951111594835916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,256,128,0.02309777836004893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,65536,0.07130310932795207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,64,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,256,32,0.012228444218635559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,51200,0.05697866943147448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,65536,1.3682550854153102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,51200,1.0721804300944011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,16384,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,65536,2.0334080590142145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,16384,0.3520506752861871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,12288,0.013815110756291283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,12288,0.2670773400200738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,16384,0.526092423333062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,10240,0.012433778080675336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,51200,1.5947866439819336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,8192,0.012381333443853589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,12288,0.39995021290249294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,10240,0.2244666682349311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,7168,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,10240,0.3370808760325114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,8192,0.18165689044528535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,6144,0.010774222513039907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,8192,0.2726293404897054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,7168,0.15816444820827907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,5120,0.00943199959066179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,7168,0.24113066991170248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,6144,0.13415911462571886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,4096,0.008409777449236976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,6144,0.20494933923085532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,5120,0.1088142196337382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,5120,0.1690151161617703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,3584,0.007663999994595845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,4096,0.08555377854241265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,3072,0.0069431112044387394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,4096,0.135644449128045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,3584,0.07481688923305936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,2560,0.00598933340774642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,3584,0.12068266338772243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,3072,0.06534222099516127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,3072,0.1059831115934584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,2048,0.0053084443012873335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,2560,0.054008887873755566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,1536,0.004879999905824661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,2560,0.09133955505159165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,2048,0.07719377676645915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,2048,0.0421706669860416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,1024,0.004267555558019214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,1536,0.033462223079469465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,768,0.003948444293604957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,1536,0.060679998662736684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,1024,0.04732622371779548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,1024,0.02553422252337138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,512,0.003843555433882607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,768,0.03978044456905789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,256,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,512,0.03392444385422601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,512,0.01759199963675605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,128,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,256,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,256,0.026836444934209187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,64,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,128,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,3072,128,128,0.022397332721286353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,128,32,0.003277333246337043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,64,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,65536,0.07013599740134345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,51200,0.05494577685991923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,32,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,16384,0.01680799987581041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,128,768,0.021293333835071985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,12288,0.015122666954994202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,16384,0.3517848915523953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,12288,0.26670755280388725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,10240,0.013436444103717804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,10240,0.22434843911064994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,51200,1.0717378192477756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,8192,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,7168,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,65536,1.368136935763889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,6144,0.00981777740849389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,7168,0.158461332321167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,8192,0.18081333902147081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,5120,0.008750222623348236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,4096,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,6144,0.13381866614023843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,3584,0.006913777854707506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,4096,0.08451466427909003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,5120,0.10872444179322983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,3072,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,2560,0.005623999983072281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,3584,0.07515911261240642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,2048,0.0052391112678580815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,3072,0.06445155541102092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,2560,0.05371288789643181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,1536,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,2048,0.04262755645645989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,1024,0.003944888710975647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,768,0.004023111114899318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,1536,0.032969776127073504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,512,0.003608889049953885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,1024,0.02549422283967336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,256,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,512,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,768,0.021191999316215515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,128,0.003141333245568805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,64,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,256,0.013517333401574029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,128,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,64,32,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,65536,0.06693689028422038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,32,0.011526222030321756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,51200,0.0521048903465271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,64,64,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,16384,0.01368711143732071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,65536,1.367802619934082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,51200,1.0713653564453125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,12288,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,10240,0.009342222577995723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,16384,0.3513920042249892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,8192,0.008660444782839881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,12288,0.2670124371846517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,10240,0.2248586548699273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,7168,0.008356444537639618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,6144,0.007720000214046902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,7168,0.15829688972897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,5120,0.007263110743628607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,6144,0.13388710551791722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,4096,0.008433777425024245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,5120,0.10895466804504395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,3584,0.007995555798212687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,4096,0.08499022324879964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,3072,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,8192,0.18054666784074572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,3584,0.07515555620193481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,2560,0.006329777754015393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,2048,0.004919111314747068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,3072,0.06454044580459595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,1536,0.004563555535342959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,2560,0.05415111117892795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,2048,0.04239911172125074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,1024,0.0041751112374994485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,768,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,512,0.003306666596068276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,1536,0.033048888047536217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,1024,0.025444444682863023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,256,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,768,0.021341333786646526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,512,0.017352888981501263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,64,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3072,32,32,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,256,0.013555555707878537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,64,0.01148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,32,0.011104889214038849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3072,32,128,0.012253333297040729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,16384,2.912659539116753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,16384,4.129160987006293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,12288,2.1585511101616754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,10240,1.7995378706190321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,12288,2.8551724751790366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,8192,1.4428311453925238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,51200,9.02407921685113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,10240,2.32041507297092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,7168,1.3249439663357205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,8192,1.7905332777235243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,7168,1.56586668226454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,6144,1.1400000254313152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,5120,0.9218826293945312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,6144,1.3284631305270724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,4096,0.7353679868910047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,4096,0.8739422162373861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,3584,0.6560408804151747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,5120,1.1067386203342013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,3072,0.5716693136427138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,3584,0.7676213582356771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,51200,14.248045179578993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,2560,0.4614017804463704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,3072,0.6628888977898492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,2048,0.3723582161797418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,1536,0.2701511118147108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,2048,0.46066490809122723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,1024,0.19543111324310303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,1536,0.362751113043891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,2560,0.561590247684055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,768,0.15135110749138728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,1024,0.2754177782270643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,512,0.11365244123670791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,256,0.08504533105426365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,768,0.21766489081912568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,512,0.18701956007215711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,256,0.16549866729312473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,64,0.06105955441792806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,128,0.15761866834428576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,32,0.06471200121773614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,64,0.15914044115278456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,65536,32,0.16202489535013834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,65536,128,0.07299377520879109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,51200,7.045729743109809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,65536,8.98254648844401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,16384,2.310741424560547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,12288,1.6959661907619898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,16384,3.286876466539171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,10240,1.4003271526760523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,51200,11.337793138292101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,12288,2.2678534189860025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,65536,14.31427001953125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,8192,1.1196319792005751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,7168,1.01018132103814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,8192,1.4440800348917644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,10240,1.8581725226508246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,6144,0.8541617923312717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,5120,0.7154657575819227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,7168,1.253417756822374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,6144,1.058916409810384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,4096,0.5932302474975586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,4096,0.6968035697937012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,3584,0.5213422245449489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,5120,0.8855813344319662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,3072,0.4124240080515544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,3584,0.6137475437588161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,2560,0.3463111188676622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,3072,0.5439004368252224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,2048,0.2790444427066379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,2560,0.4507466422186957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,1536,0.22042577796512178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,1024,0.15074311362372503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,1536,0.29222223493787974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,2048,0.37012444602118594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,512,0.09170933564503987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,1024,0.22576623492770723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,256,0.06967644559012519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,768,0.17522754934098986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,128,0.05771378013822767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,512,0.1503866646024916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,64,0.05120799938837687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,256,0.13269333044687906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,32,0.05411022239261203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,128,0.12624177667829725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,64,0.1276115576426188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,51200,32,0.12925867239634195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,51200,768,0.12184445063273112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,51200,2.3812168969048395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,65536,3.009508344862196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,16384,0.7526951366000705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,12288,0.5633742014567057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,16384,1.220008002387153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,10240,0.499434683057997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,12288,0.8668062422010633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,65536,5.350638919406467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,8192,0.4193742275238037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,10240,0.7154000070359973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,51200,4.256479051378038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,7168,0.372012456258138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,8192,0.5669795672098795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,7168,0.4951564470926921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,6144,0.31417067845662433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,5120,0.24999467531840006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,4096,0.20080533292558458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,6144,0.4289742310841878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,3584,0.1785928938123915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,5120,0.3478391170501709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,3072,0.15426488717397055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,4096,0.27636978361341685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,2560,0.13089422384897867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,3584,0.242378658718533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,2048,0.11059021949768066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,1536,0.07454755571153429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,2560,0.18105599615308973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,1024,0.05283288823233711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,2048,0.14897333251105413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,768,0.043367111020618014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,1536,0.11833777692582871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,512,0.03249955508444045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,1024,0.09080266952514648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,768,0.07233688566419813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,512,0.06194844510820177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,256,0.02458488941192627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,128,0.021755556265513103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,256,0.05297955539491442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,64,0.018897778458065458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,128,0.04998044504059685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,3072,0.21060355504353842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,16384,32,0.019134221805466544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,64,0.0502800014283922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,16384,32,0.0506568882200453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,51200,1.6534621980455186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,65536,2.1731315188937717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,16384,0.561191134982639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,16384,0.9922907087537977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,12288,0.40834845436943906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,65536,4.358100467258029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,10240,0.34085512161254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,12288,0.7299386660257975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,51200,3.4125696818033853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,8192,0.2670195632510715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,10240,0.5835413402981228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,7168,0.23833065562778047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,6144,0.21904177135891387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,7168,0.4040737681918674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,5120,0.17272444566090903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,6144,0.34441778394911027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,8192,0.464380423227946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,3584,0.12618666225009495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,5120,0.2839217715793186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,4096,0.2258524364895291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,3072,0.11096622546513875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,2560,0.09208266602622138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,3584,0.1988417837354872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,3072,0.1729137765036689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,2048,0.07470577955245972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,1536,0.06180000305175781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,2560,0.14828888575236002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,1024,0.04208711120817396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,2048,0.12278932995266384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,768,0.032580445210138954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,1536,0.09800177812576294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,4096,0.14113599724239773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,1024,0.07481333282258776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,256,0.0199644449684355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,512,0.027824888626734417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,768,0.060136000315348305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,512,0.050676445166269936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,128,0.018725333942307364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,64,0.0158942225906584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,128,0.03946933481428359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,12288,32,0.016134222348531086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,64,0.03891999853981866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,32,0.03865777783923679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,12288,256,0.04288799895180596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,51200,1.475088013543023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,65536,1.7898551093207464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,16384,0.4896142217848036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,12288,0.35182666778564453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,16384,0.8570799827575684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,10240,0.286902215745714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,12288,0.6239528656005859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,8192,0.23867556783888075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,10240,0.5346008936564127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,7168,0.19675821728176543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,65536,3.7946675618489585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,8192,0.4108480082617865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,51200,2.998228496975369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,6144,0.17886133988698324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,5120,0.1466577715343899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,7168,0.35970756742689347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,6144,0.3061315483517117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,4096,0.12013955911000569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,3584,0.1031528910001119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,5120,0.2529866695404053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,3072,0.08698311116960313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,4096,0.20086577203538683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,2560,0.07409155368804932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,3584,0.17721954981486002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,2048,0.06079910861121284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,1536,0.0490862230459849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,3072,0.1547093391418457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,2560,0.13274933232201472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,2048,0.10987289084328546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,768,0.02678488857216305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,1536,0.0877679983774821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,1024,0.06685688760545519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,512,0.02045866681469811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,256,0.016144000821643405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,768,0.05304177933269077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,128,0.013954665925767688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,512,0.043655110730065234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,64,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,256,0.03653688894377814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,32,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,128,0.033096889654795326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,64,0.0329733325375451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,10240,32,0.03257333238919576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,10240,1024,0.034862223598692156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,51200,1.2073218027750652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,65536,1.5832604302300348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,16384,0.4144124454922146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,16384,0.7382275793287488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,12288,0.28671643469068736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,51200,2.5749138726128473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,65536,3.3112044864230685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,10240,0.24976534313625762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,8192,0.19080800480312773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,12288,0.5440675417582194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,10240,0.4533413251241048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,7168,0.16769777403937447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,6144,0.141485333442688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,8192,0.3674497869279649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,7168,0.3165964550442166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,5120,0.11681777901119655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,6144,0.2681022220187717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,4096,0.09710755613115098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,3584,0.08328355683220758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,5120,0.22067022323608398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,3072,0.07341511381997003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,4096,0.17506666978200278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,2560,0.06215022007624308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,3584,0.15565866894192165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,3072,0.13601422309875488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,2048,0.050629334317313306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,1536,0.03966844413015578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,2560,0.11659199661678737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,1024,0.02942488922013177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,2048,0.09675200117958917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,768,0.0244186669588089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,1536,0.07749066750208537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,1024,0.05797244442833794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,512,0.01787911189926995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,256,0.01423911088042789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,768,0.04629777868588766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,128,0.012144889268610211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,512,0.037751999166276716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,64,0.010674667027261523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,256,0.03087377879354689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,128,0.02857066525353326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,8192,32,0.01074577785200543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,64,0.028468443287743464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,8192,32,0.02815999918513828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,51200,0.9417075051201714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,65536,1.333304935031467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,16384,0.35334311591254336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,12288,0.25492623117234975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,16384,0.6892968813578287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,10240,0.2063973347345988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,12288,0.504841751522488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,8192,0.16884977287716338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,65536,3.0606827206081815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,10240,0.42111023267110187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,7168,0.1484560039308336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,8192,0.3364080058203803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,6144,0.135480006535848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,7168,0.29441245396931964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,51200,2.3847359551323786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,4096,0.08399822314580281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,6144,0.24996532334221733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,5120,0.20457333988613555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,3584,0.07450489203135173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,4096,0.1625279982884725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,3072,0.06478222211201985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,2560,0.05459555652406481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,3584,0.14314756128523085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,2048,0.04450755649142795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,3072,0.1265946626663208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,1536,0.035749332772360906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,2560,0.10875111156039768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,1024,0.026319111386934917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,2048,0.08999377489089966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,768,0.020620443754725985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,1536,0.07192444139056735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,512,0.016007110476493835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,5120,0.10663911369111802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,256,0.012346666720178394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,1024,0.05382400088840061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,768,0.04284088810284933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,128,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,512,0.03494844502872891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,64,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,256,0.028195556667115953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,7168,32,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,128,0.02643288837538825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,64,0.02640177806218465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,7168,32,0.026100445124838088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,51200,0.8129235373602973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,65536,1.0960480372111003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,16384,0.2776533232794868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,12288,0.21312533484564888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,16384,0.6356782383388943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,10240,0.17208355002933076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,12288,0.4799066649542914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,8192,0.13881956206427679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,51200,2.178927951388889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,10240,0.38873688379923504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,7168,0.12301777468787299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,8192,0.31082844734191895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,6144,0.11570133103264703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,65536,2.7103483412000866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,5120,0.0906684464878506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,7168,0.2721048990885417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,6144,0.23082754347059461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,4096,0.07339021894666883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,3584,0.06572977701822917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,5120,0.1890746619966295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,3072,0.05640711386998495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,4096,0.1502151091893514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,2560,0.04866577850447761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,3584,0.1328444480895996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,3072,0.11554400126139323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,2048,0.03908711009555393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,2560,0.10094844632678562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,1536,0.030654221773147583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,1024,0.023296000228987798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,2048,0.08316977818806966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,768,0.017893332574102614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,1536,0.06634755267037286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,512,0.01519555515713162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,1024,0.04931555522812737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,768,0.03953333364592658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,256,0.01148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,128,0.010330666270520952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,512,0.032242665688196816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,64,0.00905600024594201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,256,0.02608088817861345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,6144,32,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,128,0.024634665913052026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,64,0.024392889605628118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,6144,32,0.024067555864651997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,51200,0.7026124530368381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,65536,0.9180835088094076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,16384,0.24316355917188856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,65536,2.520688798692491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,12288,0.17850844065348306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,51200,1.958774142795139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,10240,0.16223377651638454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,16384,0.5740835401746962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,12288,0.4278293450673421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,8192,0.11841244167751735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,7168,0.10404800044165717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,10240,0.35662489467196995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,8192,0.28592801094055176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,6144,0.08979466888639663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,5120,0.08272444539599948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,7168,0.25035733646816677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,6144,0.21210043960147432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,4096,0.06054844458897909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,5120,0.1735288831922743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,4096,0.13824355602264404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,3584,0.12166755729251438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,2560,0.037943111525641546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,3072,0.10652177863650852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,2048,0.03183733423550924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,2560,0.09063288900587295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,1536,0.024752888414594863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,2048,0.07703289058473375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,1024,0.018361777067184448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,1536,0.060211552513970264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,768,0.015020444989204407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,1024,0.04519822200139364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,3584,0.05239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,512,0.011523555550310345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,256,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,768,0.03639466563860575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,128,0.008407110969225565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,3072,0.04749422272046407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,64,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,512,0.029575109481811523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,256,0.024012444747818842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,5120,32,0.007667555577225155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,128,0.02243999971283807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,64,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,5120,32,0.022056889202859666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,51200,0.5644364356994629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,65536,0.7329840130276151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,16384,0.1965599987241957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,12288,0.1435119973288642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,16384,0.521017763349745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,10240,0.12596000565422907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,12288,0.3950666586558024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,8192,0.0964497791396247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,51200,1.7549742592705622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,10240,0.3253671063317193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,7168,0.08752355310651992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,8192,0.2599662144978841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,6144,0.07305333349439833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,65536,2.220639122856988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,5120,0.06259555286831327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,4096,0.052144000927607216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,7168,0.22807733217875162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,6144,0.19270400206247965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,3584,0.045824001232783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,5120,0.15841155582004124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,3072,0.03847733471128676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,4096,0.12589689095815024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,2560,0.033005333609051175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,2048,0.027290667096773785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,3072,0.09644711017608643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,2560,0.08253777689403959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,1536,0.022912888063324824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,2048,0.06681422392527263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,1024,0.01606399979856279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,1536,0.05376088950369093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,768,0.013125333521101209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,1024,0.040503111150529646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,512,0.01036800030204985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,768,0.03295733200179206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,256,0.008406221866607666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,256,0.021691555778185528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,3584,0.11081333292855157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,128,0.007663110891977946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,512,0.026944888962639704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,64,0.006948444578382704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,4096,32,0.007357333269384172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,128,0.020288000504175823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,64,0.020336000455750358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,4096,32,0.01998133295112186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,51200,0.4989955690171983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,65536,0.6459626621670193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,16384,0.18997511598798963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,16384,0.5153208838568794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,12288,0.1262808905707465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,12288,0.37054400973849827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,65536,2.070769839816623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,51200,1.6445066663953993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,10240,0.11307199796040852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,8192,0.09167822202046712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,7168,0.08107911215888129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,10240,0.3100035455491808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,8192,0.24809598922729492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,6144,0.06994489166471693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,5120,0.05971733066770765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,7168,0.21705599625905356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,4096,0.04741422335306803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,6144,0.18391288651360405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,3584,0.04192533426814609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,5120,0.150945782661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,4096,0.12028622627258301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,3072,0.037440889411502414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,3584,0.10566666391160752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,2560,0.03127110997835795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,2048,0.0255724440018336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,3072,0.09277688794665867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,2560,0.07859022087521024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,2048,0.06352444489796956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,1024,0.01589777734544542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,1536,0.050844444168938525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,768,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,1024,0.0380675560898251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,768,0.03151911165979173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,512,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,256,0.008406221866607666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,512,0.02533955540921953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,128,0.006962666908899943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,256,0.02094488839308421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,64,0.0063573333124319715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,32,0.006597333484225803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,128,0.01924977699915568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3584,1536,0.02183733383814494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,64,0.019222221440739103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3584,32,0.019052444232834708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,51200,0.40906132592095273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,65536,0.5292168723212348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,16384,0.1400417751736111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,12288,0.11311199929979111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,16384,0.46942488352457684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,10240,0.09049777852164374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,12288,0.3514595561557346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,8192,0.07610044214460585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,10240,0.29395556449890137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,7168,0.06662222411897448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,51200,1.540952894422743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,6144,0.06241866615083483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,8192,0.23605065875583223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,65536,1.9803536732991536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,5120,0.047820445564058095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,7168,0.20577511522505018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,4096,0.038514667087131076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,3584,0.0351493325498369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,6144,0.17464178138309053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,5120,0.14317333698272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,3072,0.029442667961120605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,4096,0.11413333151075576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,2560,0.025268443756633337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,2048,0.02184088859293196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,3584,0.10131555795669556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,3072,0.08836888604693943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,1536,0.01715022159947289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,2560,0.0749537746111552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,1024,0.013247110777431063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,2048,0.05968799855973986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,768,0.010876444478829702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,512,0.00846400029129452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,1536,0.047447111871507436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,1024,0.03584711088074578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,256,0.00703288863102595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,768,0.029493331909179688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,512,0.024479111035664875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,256,0.01958044370015462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,128,0.018227555685573153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,64,0.0060159998635451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,32,0.006246222390068903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,64,0.018026666508780587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,3072,32,0.017895999881956313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,65536,0.4522453414069281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,3072,128,0.00700711127784517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,51200,0.360609769821167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,16384,0.12573866049448648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,16384,0.4384186797671848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,12288,0.09432266818152533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,65536,1.788711971706814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,10240,0.07992533180448744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,12288,0.331059561835395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,51200,1.395435545179579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,8192,0.06124889188342624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,7168,0.05175911055670845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,10240,0.27803466055128306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,6144,0.049955556790033974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,8192,0.22340443399217394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,5120,0.038733333349227905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,7168,0.19492355982462564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,4096,0.030955556366178725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,6144,0.16424532731374106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,5120,0.13539732827080622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,3584,0.028684443897671167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,3072,0.024671110841963027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,4096,0.10793866713841756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,2560,0.021204445097181533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,3584,0.09521689017613728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,2048,0.01795288920402527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,2560,0.06975466675228544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,1536,0.014448000325096978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,2048,0.05789155430263943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,1024,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,1536,0.044943110810385815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,768,0.008819555242856344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,1024,0.033358222908443876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,512,0.007344889144102733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,768,0.027822222974565294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,256,0.005939555664857228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,512,0.02277155551645491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,128,0.005599110904667113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,256,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,64,0.004921777794758479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,128,0.016922665966881644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,3072,0.08214666446050008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2560,32,0.005325333525737126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,64,0.016901334126790363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2560,32,0.01683999929163191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,65536,0.3594382339053684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,51200,0.28166045082939994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,16384,0.09971111350589329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,16384,0.41568533579508465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,12288,0.07103821966383192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,12288,0.3144871128929986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,51200,1.2942764494154189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,65536,1.6521964603000219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,10240,0.06453600194719103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,8192,0.05398755603366428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,7168,0.049439999792310924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,10240,0.2639182143741184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,8192,0.21132266521453857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,6144,0.039573334985309176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,5120,0.03330311179161072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,7168,0.18495466974046496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,4096,0.026502221822738647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,6144,0.15563644303215876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,3584,0.0240346673462126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,5120,0.12770222293006048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,3072,0.021407110823525324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,4096,0.10185333093007405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,2560,0.018228444788191054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,3584,0.08989689085218643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,3072,0.07880355252159967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,2560,0.06721778048409356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,1536,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,2048,0.0539644459883372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,1024,0.009396444592210982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,1536,0.042465776205062866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,768,0.008014221986134848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,1024,0.031276444594065346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,512,0.006673777682913675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,768,0.026424888107511733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,256,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,512,0.021713778376579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,128,0.005548444473081165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,256,0.01756088932355245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,64,0.004983110974232356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,128,0.01591377788119846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,2048,0.015598222613334656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,2048,32,0.005264889034960005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,64,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,2048,32,0.015832889411184523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,65536,0.27522312270270455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,51200,0.2168053388595581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,16384,0.07765422264734904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,16384,0.39305867089165586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,12288,0.057915555106268994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,51200,1.214106665717231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,65536,1.5466719733344183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,10240,0.050104889604780406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,12288,0.2975635528564453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,8192,0.038076443804634943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,10240,0.24966666433546278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,7168,0.03622488843070136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,6144,0.029670221938027277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,8192,0.20047822263505724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,5120,0.026025777061780293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,7168,0.17462755574120414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,4096,0.020957332518365648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,5120,0.12052178382873535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,3584,0.019489778412712943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,6144,0.14786222245958117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,4096,0.09621422158347236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,3072,0.017786666750907898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,2560,0.01590133375591702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,3584,0.08581599924299453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,3072,0.07419200075997247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,2048,0.012484444512261285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,1536,0.010405333505736457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,2048,0.051128000020980835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,2560,0.06337866518232557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,1536,0.0398613346947564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,768,0.007384888827800751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,768,0.025099555651346844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,512,0.006334222025341458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,512,0.020706666840447318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,256,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,256,0.016232000456915963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,128,0.0046728890803125174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,1024,0.0083137775460879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,64,0.004218666503826777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,128,0.014922666880819531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1536,32,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,1024,0.02993511160214742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,64,0.014900444282425774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1536,32,0.01479555500878228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,51200,0.15458222230275473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,51200,1.1145706176757812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,65536,1.4217270745171442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,65536,0.1965937746895684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,16384,0.05083910955323113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,12288,0.044237332211600415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,10240,0.03696088989575704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,12288,0.2769928773244222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,16384,0.3646879990895589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,8192,0.02648711038960351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,7168,0.02362222141689724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,10240,0.23274755477905273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,6144,0.02144355575243632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,5120,0.01884711119863722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,8192,0.18634400102827284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,4096,0.01552088889810774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,7168,0.16303733984629312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,3584,0.014131555954615274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,4096,0.08868266476525201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,5120,0.11115645037757026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,6144,0.13706666893429226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,3072,0.012336889074908363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,3584,0.07898133330874972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,2560,0.010779555473062726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,2048,0.009467555416954888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,1536,0.00830666638082928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,3072,0.06846133205625746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,2560,0.05795911285612318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,1024,0.006641777853171031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,768,0.005920889063013925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,2048,0.04662933283381992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,1536,0.03603377938270569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,512,0.004935111022657818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,256,0.004585777719815572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,1024,0.027485332555241052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,128,0.0038559999730851916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,512,0.019304000669055514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,64,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,256,0.01552177800072564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,128,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,1024,32,0.003921777837806278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,64,0.013765333427323235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,32,0.013783111340469785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,65536,0.1853919956419203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,51200,0.14570044146643743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,16384,0.05003199974695841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,16384,0.3543733225928412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,1024,768,0.022988445229000513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,12288,0.03884889019860162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,10240,0.0325271122985416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,51200,1.0820212894015842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,12288,0.26929956012301975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,8192,0.026418666044871014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,65536,1.376400841606988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,10240,0.2255706787109375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,7168,0.022693332698610093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,6144,0.01973511113060845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,5120,0.01700533264213138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,8192,0.18092711766560873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,7168,0.15832444032033285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,4096,0.01459377838505639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,6144,0.13334222634633383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,5120,0.10858844386206733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,3584,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,3072,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,4096,0.08638577991061741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,2560,0.010427555276287926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,3584,0.07631911171807183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,2048,0.008665777742862701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,2560,0.05562577644983927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,1536,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,2048,0.04485333296987745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,3072,0.0668826699256897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,1536,0.035007112556033664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,768,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,1024,0.026879999372694228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,512,0.004990222139490975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,512,0.018952000472280715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,256,0.004304888761705822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,1024,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,128,0.00388355553150177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,256,0.014985778265529208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,128,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,64,0.003662222375472387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,768,0.022516444325447083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,768,32,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,64,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,65536,0.12671911716461182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,768,32,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,51200,0.10098044739829169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,16384,0.035747554567125105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,16384,0.3446826669904921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,12288,0.025811556312772963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,51200,1.0459796057807076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,10240,0.021801778011851843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,65536,1.3352444966634114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,8192,0.017885333961910672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,12288,0.2613582346174452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,10240,0.21940710809495714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,7168,0.015967999895413715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,6144,0.014490667316648694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,8192,0.1760880019929674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,5120,0.011876444849703046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,7168,0.15415999624464247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,4096,0.010480889015727572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,5120,0.10550489028294881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,6144,0.12989066706763372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,3584,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,3072,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,4096,0.08357777860429551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,2560,0.008003555238246918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,3072,0.0640373362435235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,2048,0.006937777830494775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,2560,0.05377599928114149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,1536,0.006401777681377199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,2048,0.04232888751559787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,1024,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,1536,0.033175110816955566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,768,0.004929777648713854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,1024,0.02573866645495097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,512,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,768,0.021712000171343487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,256,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,128,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,512,0.018227555685573153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,64,0.003269333392381668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,128,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,512,32,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,64,0.01275466630856196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,3584,0.07444622119267781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,32,0.01240533341964086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,65536,0.08563911252551609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,51200,0.06808444526460436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,16384,0.02784088916248745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,16384,0.33406133121914333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,12288,0.021711111068725586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,512,256,0.014236445228258768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,10240,0.01886399918132358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,51200,1.0146275626288521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,12288,0.2533217800988091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,8192,0.013234666652149625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,7168,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,65536,1.2955617904663086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,10240,0.2126364443037245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,6144,0.010826666322019366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,8192,0.17115022076500785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,5120,0.009386666946940953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,7168,0.14872445000542536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,4096,0.008389333056079017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,6144,0.12546577718522814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,5120,0.10213422112994724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,4096,0.08091466956668429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,3072,0.0070257774657673305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,2560,0.006717333363162146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,3584,0.07136088609695435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,2048,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,3072,0.06153421931796604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,2560,0.05146400133768717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,1536,0.0052328887912962176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,1024,0.004548444516128964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,2048,0.04208977686034309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,1536,0.03202044301562839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,768,0.004287111262480418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,1024,0.02506133251720005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,512,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,768,0.021215111017227173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,256,0.003622222277853224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,512,0.01788444485929277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,128,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,3584,0.007662222617202335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,64,0.0030008889734745026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,256,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,256,32,0.0032506665835777917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,128,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,65536,0.06857422325346205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,32,0.011889778077602386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,256,64,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,51200,0.05352088809013367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,16384,0.017408889200952318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,16384,0.3248666657341851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,12288,0.014853333433469137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,12288,0.2465075651804606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,65536,1.258853382534451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,10240,0.013316444224781461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,8192,0.012506666282812754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,10240,0.20713510778215197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,8192,0.16662311553955078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,7168,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,6144,0.010041777458455827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,7168,0.1453848944769965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,5120,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,51200,0.9867946836683484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,5120,0.09900355339050293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,4096,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,3584,0.0070159998204973005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,4096,0.07834221919377644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,3072,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,3584,0.06957155466079712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,2560,0.006068444500366847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,6144,0.12164799372355144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,2048,0.005622222191757626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,2560,0.04959200157059563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,1536,0.00490044429898262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,2048,0.03948444459173415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,1024,0.02444355520937178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,1536,0.03165422214402093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,768,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,512,0.0035866668654812705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,3072,0.05988888608084785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,256,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,512,0.017261332935757108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,768,0.020827554994159274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,128,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,64,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,256,0.013535111314720578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,128,32,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,128,0.012182222472296821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,32,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,65536,0.06596355305777656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,51200,0.050352000527911715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,16384,0.013259555730554791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,16384,0.32462933328416615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,12288,0.013209777573744455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,128,64,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,51200,0.9856177435980903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,10240,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,8192,0.010392888552612728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,65536,1.2596026526557076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,7168,0.009702222214804756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,12288,0.24634401003519693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,10240,0.2068479988310072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,8192,0.1663244432873196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,6144,0.009098666409651438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,5120,0.008633777499198914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,4096,0.007268444531493717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,6144,0.12138488557603623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,3584,0.0069520001610120135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,4096,0.07798310783174303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,3072,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,3584,0.06875199741787381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,2560,0.005912888795137405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,3072,0.05929333633846707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,2048,0.0052177777720822235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,2560,0.04935466580920749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,1536,0.004913777940803104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,2048,0.03872977693875631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,1024,0.003922666526503033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,1536,0.031130668189790513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,768,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,1024,0.02421155571937561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,7168,0.14516177442338732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,512,0.003371555565132035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,256,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,768,0.020643555455737643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,512,0.01719111038578881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,128,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,5120,0.09881422254774307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,64,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,256,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,64,32,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,128,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,32,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,65536,0.06279466549555461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,51200,0.0503039989206526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,51200,0.9863031175401477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,16384,0.012159999873903064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,64,64,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,65536,1.2581582599216037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,12288,0.010757333702511258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,10240,0.010128888818952773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,16384,0.32461243205600315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,12288,0.2464080121782091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,8192,0.008685333033402761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,7168,0.008741333252853816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,10240,0.2071039941575792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,8192,0.16635110643174914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,6144,0.007669332954618666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,5120,0.010092444717884064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,7168,0.14501155747307673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,4096,0.009511999785900116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,6144,0.12121600574917263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,3584,0.008546666966544257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,4096,0.07808178000979953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,5120,0.09881777895821466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,3072,0.007965332931942409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,2560,0.0069288888739215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,3584,0.06874933507707384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,2048,0.004953777624501122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,3072,0.059939556651645236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,2560,0.04956444435649448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,1536,0.004612444589535396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,2048,0.03864088985655043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,1024,0.004344888859324985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,768,0.003678222083383136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,1536,0.031168888012568157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,512,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,1024,0.024077332682079736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,256,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,512,0.01723288827472263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,768,0.020693333612547982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,128,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,256,0.013154666456911298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,64,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2817,32,32,0.0029306666304667792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,128,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,32,0.011376889215575324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2817,32,64,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,16384,2.7968675825330944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,51200,8.57052273220486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,16384,4.003850724962023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,12288,2.107384787665473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,16384,2.9872623019748263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,12288,2.7798292371961804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,51200,13.810820685492622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,51200,9.422809176974827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,10240,1.9166274600558812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,10240,2.274790234035916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,12288,2.220004399617513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,8192,1.4855208926730687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,8192,1.7647342681884766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,7168,1.316303147210015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,10240,1.8504871792263453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,8192,1.4836506313747828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,7168,1.5304764641655817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,6144,1.150503158569336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,7168,1.3089573118421767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,5120,0.9041119681464301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,6144,1.3007262547810872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,6144,1.1233421961466472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,4096,0.7206400235493978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,5120,1.0759564505683052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,4096,0.8525288899739584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,5120,0.9463858074612088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,3584,0.6578755378723145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,4096,0.771391126844618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,3072,0.5448560184902614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,3584,0.7494426833258735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,3584,0.6814035309685601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,3072,0.6477937698364258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,2560,0.4502284261915419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,3072,0.599689801534017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,2048,0.36533334520128036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,2560,0.5529742240905762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,1536,0.29154221216837567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,2560,0.5130835639105903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,2048,0.4532337718539768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,2048,0.4274391068352594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,1536,0.3568337758382161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,1024,0.21090932687123617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,768,0.15812799665662977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,1536,0.3408053451114231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,1024,0.27156000667148167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,1024,0.2608693440755208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,512,0.1425831052992079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,768,0.2266062100728353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,768,0.21617421838972303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,256,0.11361332734425862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,512,0.1835146612591214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,512,0.17586044470469156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,128,0.07027022043863933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,256,0.1381937795215183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,128,0.15394488970438638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,256,0.16328356001112196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,65536,128,0.1089022225803799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,32,0.06471911403867933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,64,0.15651378366682264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,65536,32,0.1584142181608412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,65536,64,0.06202666627036201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,65536,8.583363003200954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,51200,6.640706804063584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,65536,14.014686584472656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,65536,9.82017347547743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,16384,2.1591466267903647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,16384,3.1852027045355906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,51200,11.105802747938368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,16384,2.3835663265652127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,12288,1.5942470762464735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,10240,1.323854234483507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,12288,2.2388843960232205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,12288,1.8147697448730469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,10240,1.817641788058811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,8192,1.199310196770562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,10240,1.4971307118733723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,51200,7.525394863552517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,8192,1.412527084350586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,7168,0.9562373691134982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,8192,1.2071200476752388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,6144,0.8261200057135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,7168,1.2263377507527669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,7168,1.0614595413208008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,5120,0.6658942434522841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,6144,1.0358924865722656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,4096,0.5683937602572972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,6144,0.9150489171346029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,5120,0.8627377616034614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,5120,0.7676995595296224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,3584,0.4400497807396783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,4096,0.7010835541619195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,4096,0.6286755667792426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,3072,0.39987821049160427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,3584,0.601369751824273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,3584,0.5560133192274306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,2560,0.3185075653923882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,3072,0.5202070871988932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,3072,0.4889973534478082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,2560,0.4445981979370117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,2048,0.25825156105889213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,2560,0.41570043563842773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,2048,0.3639262252383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,1536,0.19942666424645317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,1024,0.14941333399878606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,2048,0.3494844436645508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,1536,0.2868604395124647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,1536,0.2798275682661268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,768,0.12121689319610596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,1024,0.21778221925099692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,512,0.08522844314575195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,768,0.17265689373016357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,768,0.176527116033766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,256,0.06643199920654297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,512,0.14519555038876003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,512,0.14819467067718506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,256,0.1307342184914483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,128,0.05579466952217949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,256,0.11339467101626927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,64,0.048700445228152804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,128,0.12386488914489746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,128,0.09127999676598443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,51200,32,0.052037331793043345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,64,0.12479200628068711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,51200,1024,0.21429155932532418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,51200,32,0.12727377149793836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,51200,2.2904239230685763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,65536,2.902919981214735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,65536,5.308202531602648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,65536,4.245369805230035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,16384,0.6954213248358833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,51200,4.174565209282769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,16384,1.1900160047743056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,12288,0.5538471009996202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,51200,3.3203697204589844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,16384,1.0792000028822157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,10240,0.5016666518317329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,12288,0.851462205251058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,10240,0.703496880001492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,8192,0.36745688650343156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,12288,0.813459555308024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,7168,0.3516275617811415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,8192,0.5544515715705024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,10240,0.6830319828457303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,8192,0.5550328890482584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,6144,0.3027084403567844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,7168,0.48598220613267684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,7168,0.48647202385796445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,5120,0.24749777052137586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,6144,0.41230400403340656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,6144,0.4173662132687039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,4096,0.1840560038884481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,5120,0.3399306668175592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,3584,0.1656079954571194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,5120,0.3512684504191081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,4096,0.27156978183322483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,4096,0.28593243492974174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,3072,0.13974222871992323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,3584,0.25293156835767955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,3584,0.238571564356486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,2560,0.11744533644782172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,3072,0.22258223427666557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,3072,0.2080328861872355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,2560,0.1788417763180203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,2048,0.09978310929404365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,2560,0.19130578305986193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,1536,0.07961333460277982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,2048,0.16103733910454646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,2048,0.1479111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,1536,0.11688444349500869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,1024,0.05432000093989902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,1536,0.13043377134535047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,768,0.04638400011592441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,1024,0.08930310938093398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,1024,0.1015662219789293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,512,0.033709334002600774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,768,0.07154933611551921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,768,0.08451110786861843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,256,0.024689777029885188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,512,0.07091555330488417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,512,0.06147999895943535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,128,0.021941334009170532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,256,0.0557804438802931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,256,0.05226755473348829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,64,0.02383199996418423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,16384,128,0.04606044292449951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,128,0.049228443039788135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,64,0.0495013329717848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,16384,32,0.05038933290375603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,16384,32,0.02407022151682112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,65536,2.279905743069119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,51200,1.7279822031656902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,65536,4.222727881537543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,16384,0.5682337548997667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,65536,3.593417697482639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,16384,0.959220462375217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,51200,3.350043614705404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,12288,0.42941776911417645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,12288,0.6941804356045194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,16384,0.9159235424465604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,10240,0.3308906555175781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,51200,2.810273700290256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,12288,0.6931155522664388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,10240,0.5740470886230469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,8192,0.2947057882944743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,7168,0.2386577659183078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,8192,0.45632266998291016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,10240,0.5901750988430446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,6144,0.21153155962626138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,8192,0.4734969139099121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,7168,0.40106932322184247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,7168,0.41465067863464355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,5120,0.17975733015272352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,6144,0.3740195433298747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,6144,0.3557599915398492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,4096,0.14419110616048178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,5120,0.2970835632748074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,5120,0.2805279890696208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,4096,0.22260000970628527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,3584,0.12285244464874268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,4096,0.24267376793755424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,3072,0.09859200318654378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,3584,0.21565777725643584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,3584,0.19646933343675402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,3072,0.17080888483259413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,2560,0.08540178007549709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,2048,0.06835999753740099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,2560,0.1630328893661499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,2048,0.12185688813527425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,1536,0.05489155650138855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,2048,0.13732977708180746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,1536,0.09677688943015204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,1536,0.1110888918240865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,1024,0.07363644573423597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,3072,0.19045066833496094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,768,0.0309840010272132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,1024,0.08707911438412136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,2560,0.1475208865271674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,768,0.05930133660634359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,512,0.022999111149046157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,768,0.07316444317499797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,256,0.01812622282240126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,512,0.05010133319430881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,512,0.062018666002485484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,128,0.015260444747077094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,1024,0.038888000779681735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,128,0.04100799891683791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,64,0.013817778064144982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,128,0.038541333542929754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,12288,32,0.014348444011476306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,64,0.0379342238108317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,32,0.0379617777135637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,12288,256,0.041913777589797974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,12288,256,0.048678222629759044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,65536,1.6838711632622614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,51200,1.3082124922010634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,65536,3.7173029581705728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,65536,3.3310854170057507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,16384,0.4723226759168837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,51200,2.935959921942817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,16384,0.8437609142727323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,12288,0.3572737905714247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,16384,0.8500800132751465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,12288,0.6325830883449978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,51200,2.6040265825059676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,10240,0.29730576939053005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,8192,0.2171795500649346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,10240,0.5086266729566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,12288,0.6418462329440647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,10240,0.5412986543443468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,8192,0.41483378410339355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,7168,0.2100213368733724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,6144,0.1777848932478163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,8192,0.4378959867689345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,7168,0.38507644335428876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,6144,0.3020382192399767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,5120,0.13688355022006565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,6144,0.3297946718004015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,4096,0.11738399664560954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,5120,0.2484497759077284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,5120,0.27593154377407497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,3584,0.09803999794854058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,4096,0.19807110892401802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,4096,0.22421601083543566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,3072,0.08822488784790039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,3584,0.17527377605438232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,7168,0.3556453386942546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,3584,0.19933777385287813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,2560,0.07340266969468859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,3072,0.15259467230902776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,3072,0.17584444416893852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,2048,0.05872088670730591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,2560,0.13153155644734701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,1536,0.046203556987974376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,2560,0.15208089351654053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,2048,0.10901066992017959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,2048,0.12722844547695583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,1024,0.03283910950024923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,1536,0.08684533172183567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,768,0.024854222933451336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,1536,0.10405066940519546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,1024,0.06605333089828491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,1024,0.08105599880218506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,512,0.019329778022236295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,768,0.0524888899591234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,768,0.06865244441562228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,256,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,512,0.0435751113626692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,512,0.05743555227915446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,128,0.0140346669488483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,256,0.04522755410936144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,256,0.03565333286921183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,64,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,128,0.03286666671435038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,10240,128,0.03679377834002177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,10240,32,0.012050666742854647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,32,0.03226666649182638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,10240,64,0.03266311022970412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,65536,1.5317209031846788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,51200,1.2067119810316298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,65536,3.2528194851345487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,16384,0.405823998981052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,65536,3.0626854366726346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,51200,2.5486897362603083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,12288,0.31173067622714573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,16384,0.7312329080369738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,51200,2.3989137013753257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,10240,0.2395519945356581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,12288,0.5366471078660753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,16384,0.7829573419358996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,12288,0.593366199069553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,8192,0.20410577456156412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,10240,0.4465991126166449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,8192,0.35530755254957413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,7168,0.17537332905663383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,8192,0.40730932023790145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,6144,0.15315732691023085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,7168,0.3110542297363281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,7168,0.3544177744123671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,6144,0.2654319869147407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,5120,0.12410577138264973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,6144,0.30392445458306205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,4096,0.10232355859544541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,5120,0.21774755583869088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,5120,0.25376976860894096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,10240,0.4997359911600749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,3584,0.08695022265116374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,4096,0.1727208826276991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,4096,0.2065404388639662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,3072,0.07358666923311022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,3584,0.15378044711218938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,2560,0.06314844555324979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,3072,0.13432266977098253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,2560,0.11577688323126899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,2560,0.13909155792660183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,2048,0.05236977669927809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,2048,0.09603022204505073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,2048,0.11768533123864068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,1536,0.040229333771599665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,3584,0.18437777625189888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,3072,0.1617386606004503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,1024,0.029768887493345473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,1536,0.07639733288023207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,1536,0.0959013303120931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,768,0.023792889383104112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,1024,0.05745066536797417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,1024,0.07523822122149997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,512,0.018065777089860704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,768,0.06281066603130765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,768,0.045872890286975436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,256,0.014160888062583076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,512,0.03749600052833557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,512,0.05361244413587782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,128,0.012505778008037143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,256,0.04154488775465224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,256,0.030299554268519085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,64,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,128,0.02847733431392246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,8192,32,0.013119999733236102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,8192,128,0.0336168905099233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,64,0.028549333413441975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,8192,32,0.027793778313530818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,65536,1.490583101908366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,51200,1.134882715013292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,65536,2.943112055460612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,65536,2.937880833943685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,16384,0.3826631175147162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,51200,2.334718280368381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,16384,0.6776568624708387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,12288,0.28440978791978627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,16384,0.7506586710611979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,12288,0.498327096303304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,10240,0.23299111260308158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,51200,2.297638151380751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,12288,0.5681742032368978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,10240,0.4142186641693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,8192,0.18268267313639322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,10240,0.5201573371887207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,7168,0.15505954954359266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,8192,0.33095023367140025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,8192,0.3873306645287408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,6144,0.13346132967207167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,7168,0.2899911138746473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,5120,0.11107289128833348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,7168,0.33987821473015684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,6144,0.24588444497850206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,6144,0.2903093232048882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,4096,0.0909315546353658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,5120,0.20172977447509766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,5120,0.2430808809068468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,4096,0.16108266512552896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,3584,0.07921955320570204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,4096,0.19737866189744738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,3072,0.06955199771457247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,3584,0.1758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,3072,0.1255520052380032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,2560,0.059319999482896596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,3584,0.14228799608018664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,3072,0.15434222751193577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,2048,0.048123555050955884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,2560,0.1082000003920661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,2560,0.13330222500695124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,1536,0.031333333916134305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,2048,0.08955911133024429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,1024,0.023216888308525085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,2048,0.11300622092352973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,1536,0.07150933477613661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,1536,0.09125510851542155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,768,0.02030311193731096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,1024,0.05319911241531372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,1024,0.07224355803595649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,512,0.017571555243598092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,768,0.04245511028501722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,768,0.060545775625440806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,256,0.013764444324705334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,512,0.051170667012532554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,128,0.012072888513406118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,256,0.03930222325854831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,256,0.028147555059856836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,512,0.03506755497720506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,64,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,7168,32,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,128,0.026568889617919922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,7168,128,0.03193688723776076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,64,0.026341333985328674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,7168,32,0.025786666406525507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,51200,0.908225801255968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,65536,1.1212417814466689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,65536,2.7389305962456594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,16384,0.29359732733832467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,51200,2.1322222815619574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,65536,2.6747040218777127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,12288,0.22014666928185356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,51200,2.09718746609158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,16384,0.6217493481106228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,16384,0.6863839891221789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,12288,0.4595191213819716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,10240,0.20038310686747232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,12288,0.5213066736857096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,8192,0.15404533015357122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,10240,0.3844124476114909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,10240,0.44251110818650985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,8192,0.3062551021575928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,7168,0.13851911491817898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,8192,0.3564488887786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,7168,0.2678702142503527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,7168,0.3114017645517985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,6144,0.11638311545054118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,5120,0.09649599923027886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,6144,0.22752979066636828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,6144,0.2654044363233778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,4096,0.07853955692715116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,5120,0.2225360075632731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,5120,0.18749688731299508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,3584,0.06673155228296916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,4096,0.14878043863508436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,4096,0.18106667200724283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,3072,0.06040622128380669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,3584,0.16058754920959473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,3584,0.13153333134121364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,2560,0.048671109808815845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,3072,0.11460977130466038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,3072,0.1419102218416002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,2048,0.04135911001099481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,2560,0.12263822555541992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,2560,0.10013510783513387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,1536,0.03127200073666043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,2048,0.0825715528594123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,2048,0.10371733374065822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,1024,0.02474044428931342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,1536,0.08463466829723781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,1536,0.0655297769440545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,768,0.018956444329685636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,1024,0.04866666595141093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,1024,0.06694844696256849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,512,0.014174222946166992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,768,0.05615999963548449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,768,0.03975911272896661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,256,0.01130844404300054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,512,0.031874666611353554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,512,0.04629155662324694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,128,0.00996888925631841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,256,0.03671288821432326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,256,0.026165333059098985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,64,0.009053332938088311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,128,0.024078221784697637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,6144,128,0.030925333499908447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,6144,32,0.009689778089523315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,64,0.024359110328886244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,6144,32,0.023790222075250413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,65536,0.8839982350667318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,51200,0.7291830910576714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,65536,2.4456950293646917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,16384,0.22815110948350695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,51200,1.926969740125868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,65536,2.540853288438585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,12288,0.17830133438110352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,16384,0.5701040161980523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,12288,0.42202043533325195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,51200,1.9937207963731554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,10240,0.14500889513227674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,12288,0.49649688932630753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,8192,0.12027288807762994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,10240,0.35161156124538845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,10240,0.41719733344184023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,7168,0.10677778058581883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,8192,0.338642676671346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,8192,0.28224889437357586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,6144,0.08930133448706733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,16384,0.6538320117526585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,7168,0.24645688798692492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,5120,0.07308444711897108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,7168,0.29677777820163304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,6144,0.2527680132124159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,6144,0.20948533217112222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,4096,0.05847111013200548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,5120,0.21133155292934844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,3584,0.05431644452942742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,4096,0.1725484397676256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,4096,0.13735555277930364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,3072,0.04433777928352356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,3584,0.12124178144666885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,3584,0.15327110555436876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,2560,0.038337777058283486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,3072,0.13481244775984022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,3072,0.10599644316567315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,2048,0.03274044394493103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,2560,0.09044266409344143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,2560,0.11639289061228435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,1536,0.02586666742960612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,2048,0.07656444443596734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,5120,0.17133422692616782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,2048,0.09885688622792561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,1024,0.018248000078731112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,1536,0.05941510862774319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,1536,0.08117333385679457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,768,0.015107555521859063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,1024,0.04459466536839803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,1024,0.0632115536265903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,512,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,768,0.054069333606296115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,768,0.035997334453794695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,512,0.02875822120242649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,256,0.009114666945404476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,512,0.04526399903827243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,128,0.008348444269763099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,256,0.023407111565272015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,256,0.03572266631656223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,64,0.00738933351304796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,128,0.02206755512290531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,5120,128,0.030293332205878362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,5120,32,0.007700444095664554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,64,0.022131555610232886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,5120,32,0.021735111872355144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,65536,0.8358417616950141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,51200,0.6410035557217068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,65536,2.1726542578803167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,65536,2.412449730767144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,51200,1.7241040335761175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,16384,0.5165777736239964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,16384,0.22923021846347383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,16384,0.6215350892808702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,12288,0.17002577251858184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,12288,0.38634488317701554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,12288,0.47138224707709414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,10240,0.1247840060128106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,8192,0.09715111388100518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,10240,0.32227733400132924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,51200,1.8890026940239801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,10240,0.39698134528266055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,8192,0.25807200537787545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,7168,0.08694399727715386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,6144,0.07700977722803752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,8192,0.32156801223754883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,7168,0.22634310192532012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,7168,0.2813111146291097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,5120,0.06116000148985121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,6144,0.24012533823649088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,6144,0.19136177168952095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,4096,0.05356711149215698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,5120,0.15667733881208631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,5120,0.20120978355407715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,3584,0.04541688826349047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,4096,0.1632195578681098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,3584,0.11058311329947577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,3072,0.03990311092800564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,3584,0.14557422531975642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,2560,0.033257779147889875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,3072,0.09638044569227432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,3072,0.12780977620018855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,2560,0.08204799890518188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,2048,0.028852442900339764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,2560,0.1110959980222914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,1536,0.022978666755888198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,2048,0.09287555350197686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,2048,0.06562755505243938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,1024,0.015841777125994366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,1536,0.053429332044389516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,1024,0.0403128895494673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,768,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,1024,0.06036800146102905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,4096,0.12548889054192439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,512,0.010334221853150262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,768,0.050957332054773964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,512,0.026538666751649644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,512,0.04117600123087565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,256,0.008364444805516137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,256,0.02142400046189626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,128,0.007932444413503012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,256,0.033794667985704206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,1536,0.07617511351903279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,128,0.020015110572179157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,64,0.007015110717879401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,4096,32,0.006973333656787872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,4096,128,0.028498666154013738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,64,0.019903111788961623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,32,0.019679110911157396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,4096,768,0.03256266646915012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,51200,0.6708213488260905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,65536,0.8154657681783041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,65536,2.089273876614041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,16384,0.20715910858578154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,51200,1.6226773791842992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,65536,2.410355461968316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,16384,0.49135557810465497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,51200,1.8885093265109594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,12288,0.15595377816094294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,10240,0.1284959978527493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,16384,0.6209137704637315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,12288,0.3667866653866238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,8192,0.09886844290627374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,10240,0.306675566567315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,10240,0.3957804573906793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,7168,0.0864586697684394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,8192,0.24572534031338164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,8192,0.3210222191280789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,7168,0.2152257760365804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,6144,0.0744382209248013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,7168,0.28011732631259495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,6144,0.1821733315785726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,6144,0.23915200763278535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,12288,0.47164000405205625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,5120,0.0629262195693122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,4096,0.05008977651596069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,5120,0.1493591070175171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,5120,0.20009156068166098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,4096,0.11956532796223958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,3584,0.04369599951638115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,4096,0.16241955757141113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,3584,0.1055155528916253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,3584,0.14512711101108128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,3072,0.09203733338250054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,2560,0.03162755568822225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,3072,0.12756889396243626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,2560,0.07836177614000109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,2048,0.02607466611597273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,2560,0.11079021957185532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,2048,0.0631128880712721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,1536,0.0210844443904029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,2048,0.09338488843705918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,3072,0.03762400150299072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,1024,0.016287111573749118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,1536,0.05056088831689623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,1536,0.07576800054974027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,1024,0.03795555565092299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,768,0.013083555632167392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,1024,0.06000355879465739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,512,0.01000355515215132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,768,0.031239999665154353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,768,0.049560887946022876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,256,0.007924444145626491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,512,0.025831111603313025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,512,0.040468444426854454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,128,0.006973333656787872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,256,0.03244888782501221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,256,0.0207360006040997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,64,0.0063662222690052455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,128,0.018959111637539335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3584,32,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3584,128,0.028148445818159316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,64,0.01924088928434584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3584,32,0.018952000472280715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,65536,0.5478880140516493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,51200,0.43573689460754395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,65536,1.9420346154106989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,16384,0.15253778298695883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,51200,1.5342897839016383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,65536,2.279933293660482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,12288,0.11965600649515788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,16384,0.4637129041883681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,16384,0.5887857543097602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,12288,0.3481360011630588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,10240,0.09946222437752618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,12288,0.44798045688205296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,8192,0.07856177621417575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,10240,0.2915911144680447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,51200,1.7847590976291234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,10240,0.3766320016649034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,7168,0.07382488913006253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,8192,0.2338933414883084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,8192,0.30561865700615776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,6144,0.05912355581919352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,7168,0.2046977811389499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,5120,0.04980533321698507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,7168,0.2662888897789849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,6144,0.17304355568355986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,4096,0.04034844372007582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,5120,0.1416862275865343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,5120,0.1898933384153578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,3584,0.035563555028703474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,4096,0.11326666673024495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,4096,0.15485155582427979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,3072,0.032115555471844144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,3584,0.09972977638244629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,3584,0.13757867283291286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,2560,0.02712000078625149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,3072,0.08649155828687881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,3072,0.12136622269948323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,2048,0.022498667240142822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,2560,0.07358133130603366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,2560,0.1049928863843282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,1536,0.018592889110247295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,2048,0.058712886439429395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,2048,0.08819022443559434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,6144,0.2275164392259386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,1024,0.014352889524565803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,1536,0.046485334634780884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,1536,0.07227733400132921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,768,0.011343110766675739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,1024,0.05603555838267008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,1024,0.03570933474434747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,512,0.008656889200210571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,768,0.04657422171698677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,768,0.029176887538697984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,256,0.0068897778789202375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,512,0.03969600134425693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,512,0.024336000283559162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,128,0.0063475556671619415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,256,0.019341333044899833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,256,0.03174755639500088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,64,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,3072,128,0.027422222826215956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,3072,32,0.005987555616431766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,128,0.018188445104493033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,64,0.017954667409261067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,3072,32,0.017904000149832833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,65536,0.4378631114959717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,51200,0.3241413434346517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,65536,1.7816552056206598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,51200,1.3850444157918294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,16384,0.12029243840111627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,65536,2.152097702026367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,16384,0.4372417661878798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,12288,0.08432088957892524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,16384,0.5567768944634331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,51200,1.6859555774264867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,12288,0.3313208950890435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,12288,0.4234328799777561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,10240,0.08597777949439155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,8192,0.06984088818232219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,10240,0.2774702178107368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,10240,0.35631643401251895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,8192,0.22243732876247832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,7168,0.056108441617753774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,7168,0.19375199741787383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,8192,0.28963110182020396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,7168,0.2529626687367757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,6144,0.0487937761677636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,5120,0.03830577929814657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,6144,0.2154159943262736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,6144,0.16418755054473877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,5120,0.1345484389199151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,4096,0.03184533450338576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,5120,0.17975021733178032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,4096,0.10753511057959662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,3584,0.0292595558696323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,4096,0.14645332760281032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,3072,0.02535555594497257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,3584,0.09465689129299587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,3584,0.13072800636291504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,2560,0.021330666210916307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,3072,0.08330044481489393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,3072,0.11477688948313396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,2048,0.018096889058748882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,2560,0.09924533632066514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,1536,0.01514844430817498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,2048,0.08400622341367935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,2048,0.058232890235053174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,1024,0.010995555255148145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,1536,0.06843822532229953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,1024,0.03399822115898132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,768,0.009202666580677032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,1024,0.05333955420388115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,768,0.044222222434149853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,512,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,768,0.027746667464574177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,2560,0.0687075588438246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,512,0.022679999470710754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,512,0.03821511069933573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,256,0.018231110440360177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,1536,0.04484533270200094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,128,0.005538666827811136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,256,0.031288888719346784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,64,0.005000888887378904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,128,0.016887111796273124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2560,128,0.026103999879625108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,32,0.0052986666560173035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,64,0.016920889417330425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2560,32,0.016905777984195285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,65536,0.42336445384555393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2560,256,0.006242666807439592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,51200,0.33432534005906844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,65536,1.6519769032796223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,51200,1.2902498245239258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,16384,0.11419022083282471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,16384,0.41348711649576825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,65536,2.147957271999783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,12288,0.08287733130984835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,51200,1.684115515814887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,16384,0.556696891784668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,10240,0.07252444161309136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,12288,0.313281774520874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,8192,0.04907911022504171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,10240,0.2629484335581462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,12288,0.4235520097944472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,10240,0.3554195562998454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,7168,0.0457911127143436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,8192,0.21086755063798693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,6144,0.037412444750467934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,8192,0.28909511036343044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,7168,0.18322400252024332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,7168,0.2524684535132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,5120,0.03278222348954942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,6144,0.15459467305077448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,5120,0.1274968915515476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,6144,0.21471288469102648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,5120,0.17870755990346274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,4096,0.02666577696800232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,3584,0.02421955598725213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,4096,0.14496178097195095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,4096,0.10214400291442871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,3584,0.08943644497129653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,3072,0.021278222401936848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,3584,0.1295528941684299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,2560,0.018271999226676095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,3072,0.07850044303470187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,3072,0.11422044701046413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,2048,0.016191111670600045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,2560,0.06672710842556424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,2560,0.09951555728912354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,1536,0.012537777423858643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,2048,0.08344799942440456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,2048,0.055326223373413086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,1536,0.04231999980078804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,1536,0.06821511189142863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,1024,0.031173331869973078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,768,0.008026666939258575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,1024,0.052114665508270264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,768,0.02610844373703003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,768,0.04352533486154345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,512,0.006715555571847492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,512,0.021359110871950786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,512,0.03725511166784499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,256,0.005921777751710679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,256,0.017112889223628573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,256,0.029899554120169744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,1024,0.00959022260374493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,128,0.00498933345079422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,64,0.004871110949251386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,128,0.01587288909488254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,2048,128,0.02563200063175625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,64,0.015586665934986539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,2048,32,0.015912888778580558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,65536,0.28462311956617564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,2048,32,0.004889777965015835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,51200,0.2358000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,65536,1.5169297324286566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,65536,2.020999060736762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,16384,0.08317066563500299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,51200,1.1897591484917533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,16384,0.38648356331719297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,12288,0.06224355432722303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,16384,0.523439089457194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,10240,0.052601777844958834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,51200,1.584576924641927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,12288,0.29306666056315106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,8192,0.04085333479775323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,10240,0.2456293370988634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,12288,0.39859734641181094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,7168,0.03585422370168898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,8192,0.19694577323065865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,10240,0.3364364571041531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,6144,0.03131999903255039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,7168,0.17211555110083687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,7168,0.23800444602966309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,5120,0.027280888623661462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,6144,0.20219733979966906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,6144,0.1447546614540948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,4096,0.02257866660753886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,5120,0.11807110574510361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,5120,0.16840622160169813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,3584,0.020092444287406076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,4096,0.13727911313374838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,4096,0.09531555573145549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,3072,0.018135999639829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,3584,0.0834524432818095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,3584,0.12197777960035537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,2560,0.01625333395269182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,3072,0.07300711340374418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,3072,0.10727022091547649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,2048,0.013321778012646569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,2560,0.062014222145080566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,2560,0.09376266929838394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,2048,0.04996088809437222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,1536,0.011185777684052786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,2048,0.07899110847049289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,1024,0.008756444685988957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,1536,0.038712888956069946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,1536,0.06293600135379367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,1024,0.02892711096339756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,768,0.007640889121426477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,1024,0.04903911219702827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,512,0.006278222219811545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,768,0.02437155610985226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,768,0.041800889703962535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,8192,0.27241333325703937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,256,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,512,0.02027555637889438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,512,0.03561333484119839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,128,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,256,0.016216889023780823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,256,0.028840889533360798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,128,0.014905777242448596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1536,128,0.024383111132515803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,32,0.005015999906592899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,64,0.01484444406297472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1536,32,0.014562666416168213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,65536,0.198600000805325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1536,64,0.004527111020353106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,51200,0.16017956203884548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,65536,1.418906635708279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,51200,1.112884415520562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,65536,2.0203715430365667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,16384,0.05459022190835741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,16384,0.3649280071258545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,16384,0.5231111314561632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,12288,0.03949600127008226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,51200,1.5818452835083008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,10240,0.033096889654795326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,12288,0.27685244878133136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,12288,0.3991226620144314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,10240,0.23216443591647676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,8192,0.02754844393995073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,10240,0.3349608845180935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,7168,0.024392000503010217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,8192,0.18653689490424263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,8192,0.2727102173699273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,6144,0.021486222743988037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,7168,0.16173244847191706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,7168,0.2379697693718804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,5120,0.018944000204404194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,6144,0.1367031070921156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,6144,0.20206577248043486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,4096,0.015311110350820752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,5120,0.1115680005815294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,5120,0.1682880057228936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,4096,0.08898666832182144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,4096,0.13687554995218912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,3584,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,3072,0.012991111311647626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,3584,0.0787280003229777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,3584,0.12257333596547444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,2560,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,3072,0.06895555390252008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,3072,0.10703911383946736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,2048,0.009395555489593083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,2560,0.057806220319536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,2048,0.04606666829850939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,1536,0.008007999923494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,2048,0.07832177480061848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,1024,0.00675733346078131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,1536,0.03656889001528422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,1536,0.06401422288682726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,768,0.005893333504597346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,1024,0.027499554885758296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,1024,0.04863111178080241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,512,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,768,0.02299022177855174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,768,0.04125688804520501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,256,0.004598222259018156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,512,0.018943111101786297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,512,0.03529777791765001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,128,0.0038622220357259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,256,0.015201777219772339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,256,0.028232001596026953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,64,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,128,0.014164444473054675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,128,0.0244151105483373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,1024,32,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,64,0.013783999615245394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,1024,32,0.01348355578051673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,65536,0.16557244459788004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,51200,0.12879110707177058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,1024,2560,0.09385422203275894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,51200,1.078446176317003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,65536,1.3740897708468969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,16384,0.04669955703947279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,16384,0.3538675573137071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,12288,0.034111999803119235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,16384,0.49040264553493923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,65536,1.8890711466471355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,10240,0.027878221538331773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,12288,0.2690524525112576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,10240,0.22583111127217612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,12288,0.37393776575724286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,8192,0.022127999199761286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,10240,0.314210679796007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,51200,1.481923527187771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,8192,0.1812035507626004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,7168,0.019494222270117868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,8192,0.25562577777438694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,7168,0.22262843449910483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,6144,0.13313599427541098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,5120,0.015670221712854173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,6144,0.18910045093960234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,5120,0.10815999905268352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,4096,0.012864888542228274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,5120,0.15706311331854925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,4096,0.08604622549480861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,3584,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,4096,0.12842044565412733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,6144,0.017851556340853374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,3072,0.010408889088365767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,3584,0.0760657787322998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,3584,0.11409333017137314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,7168,0.15739111105600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,3072,0.06609600120120578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,2560,0.009492444495360056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,3072,0.10073688957426284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,2048,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,2560,0.0874648888905843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,2048,0.044545776314205594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,1536,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,2048,0.07405600282880995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,1536,0.03537333342764113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,1024,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,1536,0.05817511346605089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,1024,0.02665244374010298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,768,0.0052897776994440295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,1024,0.046322666936450534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,2560,0.05561155743069119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,768,0.02236622240808275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,512,0.004575110971927643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,768,0.03904622130923801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,256,0.004184000194072723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,512,0.0332897769080268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,512,0.01864622202184465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,128,0.0038693332009845306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,256,0.02716444432735443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,256,0.01493866741657257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,64,0.003616888903909259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,768,128,0.02318222158484989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,768,32,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,64,0.013510222236315409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,32,0.012882666455374824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,65536,0.13024622864193386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,768,128,0.013263999587959714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,51200,0.10299822356965806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,65536,1.3345858256022136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,51200,1.046520021226671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,16384,0.03529689047071669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,65536,1.8886426289876301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,16384,0.343905766805013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,12288,0.0262080000506507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,16384,0.49071820576985675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,51200,1.4780844582451715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,10240,0.021456888980335657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,12288,0.26109690136379665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,12288,0.3736586570739746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,8192,0.018254222141371835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,10240,0.3139466709560818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,8192,0.17635733551449248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,7168,0.01588444411754608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,8192,0.2542195585038927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,6144,0.014554666148291694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,7168,0.15307645003000894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,7168,0.22210489379035103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,5120,0.01204622205760744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,6144,0.12929956118265787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,10240,0.21898310714297822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,6144,0.18924000528123644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,4096,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,5120,0.15845599439409044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,3584,0.009894222021102905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,4096,0.08392444584104751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,4096,0.12763467099931505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,3072,0.009007111191749573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,3584,0.0736115574836731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,3584,0.11431911256578232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,3072,0.0640266670121087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,2560,0.007689778175618913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,3072,0.10061244169871013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,2048,0.0069653333889113525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,2560,0.05315022336112129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,2560,0.08717511097590129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,1536,0.00630044440428416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,5120,0.10522933138741387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,2048,0.07296888695822822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,1024,0.005233777893914117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,1536,0.033666667011049055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,1536,0.05881333351135254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,768,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,1024,0.046153777175479464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,768,0.021712000171343487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,512,0.004230222354332606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,768,0.03906311260329352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,512,0.018221333622932434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,256,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,512,0.03342933456103007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,256,0.014463111758232117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,128,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,2048,0.042951110336515635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,256,0.027225777506828308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,64,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,128,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,512,128,0.02331022255950504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,512,32,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,1024,0.025711110896534387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,65536,0.08569955825805664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,64,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,512,32,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,51200,0.06828977664311726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,51200,1.0146666632758248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,16384,0.02747022277779049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,65536,1.2948267194959853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,51200,1.471336046854655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,16384,0.333168003294203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,12288,0.020621332857343886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,65536,1.874667485555013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,12288,0.25290756755405003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,10240,0.018974221414989896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,16384,0.4874720043606228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,12288,0.3707919915517171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,8192,0.01648177703221639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,10240,0.21258310476938883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,8192,0.17075110806359184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,10240,0.3121573395199246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,7168,0.01461955573823717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,8192,0.25305599636501735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,6144,0.012545777691735161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,7168,0.22064710987938774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,7168,0.14847911728752985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,5120,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,6144,0.1250657770368788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,6144,0.1876044405831231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,4096,0.008366222182909647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,5120,0.10265866915384929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,5120,0.15562933021121556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,3584,0.007672888537247975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,4096,0.08078577783372667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,4096,0.12696533732944065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,3584,0.07133777936299641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,3072,0.007335999773608313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,3584,0.11312711238861084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,2560,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,3072,0.061787552303738065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,3072,0.09954577684402466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,2048,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,2560,0.05101422137684292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,2560,0.08550666438208686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,1536,0.005236444373925527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,2048,0.07220888800091214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,2048,0.04044177797105577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,1024,0.004590222405062782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,1536,0.031947556469175555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,1536,0.05688711007436117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,768,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,1024,0.04531022244029575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,1024,0.02514044443766276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,512,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,768,0.021049777666727703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,768,0.03846577803293864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,256,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,512,0.03293955657217238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,512,0.01812799937195248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,128,0.003264888914095031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,256,0.013771555489963956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,256,0.026759111218982275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,64,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,256,128,0.02269688910908169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,256,32,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,128,0.012494222157531314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,64,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,65536,0.06553600231806438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,256,32,0.01180266671710544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,51200,0.053921778996785484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,51200,0.9856986999511719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,65536,1.258072853088379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,16384,0.018603555030292936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,16384,0.32418399386935765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,12288,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,65536,1.8770328097873266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,16384,0.48672442966037327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,10240,0.012515555653307172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,12288,0.24678667386372885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,12288,0.3707759910159641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,8192,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,10240,0.2068328857421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,10240,0.31114400757683647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,51200,1.4690666198730469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,7168,0.010704889065689512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,8192,0.16677421993679473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,6144,0.010432889064153036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,7168,0.1447768873638577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,8192,0.2531893253326416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,7168,0.22032622496287027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,5120,0.009407111340098912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,6144,0.18731110625796846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,4096,0.008025777836640676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,5120,0.15590666400061712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,4096,0.07839288976457384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,3584,0.0075786668393347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,4096,0.126510222752889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,3584,0.06956799825032552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,3584,0.11339555846320258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,3072,0.0069546666410234236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,3072,0.06025333536995781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,3072,0.09971466991636489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,2560,0.006416000011894438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,6144,0.12141156196594238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,2048,0.005617777920431561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,2560,0.05005333489841885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,2560,0.08620178037219578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,1536,0.0047324444684717385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,2048,0.07106222046746148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,2048,0.03981688949796889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,1536,0.03156088789304098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,1024,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,1536,0.05746666590372721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,1024,0.024325332707828943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,768,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,1024,0.04528533418973287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,5120,0.09875200192133586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,512,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,768,0.020661332541041903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,768,0.03813599877887302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,256,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,512,0.01722311145729489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,512,0.032297776805029974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,256,0.02610133257177141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,256,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,64,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,128,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,128,32,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2816,128,128,0.022340445054901972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,65536,0.06576444705327351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,64,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,128,32,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,51200,0.051853332254621715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,16384,0.014112888111008538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,16384,0.32435377438863117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,12288,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,51200,0.985040028889974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,10240,0.010002666877375709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,65536,1.2577661938137479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,12288,0.2467742231157091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,8192,0.009355555805895064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,7168,0.009004444711738164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,10240,0.20684266090393066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,6144,0.00833777752187517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,8192,0.16636710696750218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,7168,0.14478311273786756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,5120,0.008081778056091731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,4096,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,6144,0.12127644485897487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,3584,0.0069448889957533936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,5120,0.09852533207999335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,3072,0.006266666783226862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,4096,0.07840800285339355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,3584,0.06907822026146783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,2560,0.005734222216738596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,2048,0.005286222116814719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,3072,0.058851553334130176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,1536,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,2560,0.04909777641296387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,1024,0.00388711111413108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,2048,0.03879377908176846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,1536,0.030993776188956365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,768,0.0039813332259655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,512,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,1024,0.024227556255128648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,256,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,768,0.02030933399995168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,128,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,512,0.017265778448846605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,256,0.01311555587583118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,64,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,128,0.011932444241311816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,64,32,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,64,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,65536,0.06236355834537082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,64,32,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,51200,0.05053244365586174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,16384,0.012788444757461548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,16384,0.32424622111850315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,12288,0.01033866653839747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,51200,0.985015975104438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,10240,0.009334222310119206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,65536,1.2580817540486653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,12288,0.24663111898634169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,8192,0.008660444782839881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,7168,0.007995555798212687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,6144,0.007657777931955125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,10240,0.20650488800472686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,5120,0.007115555306275685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,7168,0.1441191169950697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,8192,0.16608799828423396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,4096,0.008356444537639618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,6144,0.12120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,5120,0.0984462234708998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,3072,0.006960888703664144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,4096,0.07793955670462714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,3584,0.06877511077457003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,2560,0.0063591111037466265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,2048,0.005304000029961268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,3072,0.05925866630342272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,1536,0.004530666602982415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,2560,0.04919999837875366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,2048,0.039142221212387085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,1536,0.030881779061423406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,768,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,1024,0.02403733299838172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,512,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,768,0.020263110597928364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,3584,0.007666666474607255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,512,0.017186666528383892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,256,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,128,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,256,0.013150222599506378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,64,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,128,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,32,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,64,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2816,32,32,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2816,32,1024,0.004168888760937585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,16384,2.5519252353244357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,16384,3.764654371473524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,12288,1.919048097398546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,51200,7.97869873046875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,10240,1.5927653842502172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,12288,2.613636440700955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,10240,2.1154941982693143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,8192,1.2878942489624023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,8192,1.6410675048828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,7168,1.1323315302530925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,6144,0.9554071426391602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,7168,1.4237360424465602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,6144,1.212580468919542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,5120,0.8132408989800347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,51200,12.978464762369791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,5120,1.0030683941311307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,4096,0.6913910971747504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,3584,0.5851688914828831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,3072,0.49872530831231016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,4096,0.8031182289123535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,3584,0.7005048857794868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,3072,0.6202008989122179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,2560,0.4103991190592448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,2048,0.33077600267198354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,2560,0.515974203745524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,1536,0.2637351089053684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,2048,0.41927467452155215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,1024,0.18783733579847547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,1536,0.33214489618937176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,768,0.16066043906741673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,512,0.12002755535973443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,1024,0.2707395553588867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,768,0.20028889179229736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,256,0.08620978064007229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,512,0.1714480055703057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,128,0.0681413345866733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,256,0.1522515614827474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,64,0.05948355462816027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,65536,32,0.06297422117657132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,128,0.1443271107143826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,32,0.14778489536709255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,65536,64,0.14698399437798396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,51200,6.399631924099392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,65536,8.299874199761286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,16384,2.18192270067003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,51200,10.326954311794704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,12288,1.5774106979370117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,16384,2.994530571831597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,10240,1.3231386608547635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,12288,2.0823182000054254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,65536,13.075934516059029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,8192,1.0494551128811307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,10240,1.6890818277994792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,7168,0.8835511207580566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,6144,0.7917111184861926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,8192,1.3103218078613281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,7168,1.1416924794514973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,5120,0.6657084359063042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,4096,0.5101253191630045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,6144,0.9657724168565539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,5120,0.8078791300455729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,3584,0.4655511114332411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,4096,0.6479697757297093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,3072,0.38263821601867676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,3584,0.56095642513699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,2560,0.32088711526658803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,2048,0.27009688483344185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,3072,0.5182293256123861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,2560,0.4143422179751926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,1536,0.2046088907453749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,2048,0.3367226653628879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,1024,0.13774755265977648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,1536,0.26697156164381236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,768,0.1081742246945699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,512,0.08092266983456083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,256,0.0631271137131585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,512,0.13828889528910318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,1024,0.20858666631910536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,768,0.16128267182244196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,128,0.05752888653013441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,64,0.050332443581687085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,256,0.12232000297970241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,51200,32,0.05243200063705444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,128,0.11611821916368271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,64,0.11787111229366726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,51200,32,0.11815199587080215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,51200,2.1101226806640625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,65536,2.6275013817681208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,16384,0.6881448957655164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,16384,1.1087466345893013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,51200,3.863811492919922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,12288,0.504842652214898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,65536,4.915457831488715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,12288,0.7874533335367838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,10240,0.46447377734714085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,8192,0.36703912417093915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,10240,0.6521217558119032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,8192,0.5290266672770182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,7168,0.3200391133626302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,6144,0.2746391031477186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,7168,0.44816356235080296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,6144,0.3818097909291585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,5120,0.23136888609992134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,4096,0.1826711098353068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,3584,0.16065511438581678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,4096,0.2521973450978597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,5120,0.31514043278164333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,3072,0.13897600438859728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,2560,0.11797066529591878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,3584,0.22199289004007974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,3072,0.193378660413954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,2048,0.09544710980521308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,1536,0.0735422240363227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,2560,0.16670400566524932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,2048,0.13679822285970053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,1024,0.05268355541759067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,768,0.042136000262366406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,1536,0.10897955629560684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,1024,0.08347022533416748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,512,0.032467554012934365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,256,0.02455377744303809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,768,0.06726933187908597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,128,0.020647111866209242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,512,0.05752444267272949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,64,0.018187556001875136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,256,0.049327999353408813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,16384,32,0.01888177792231242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,128,0.047006222936842174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,64,0.04694222079383003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,16384,32,0.046161777443355985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,51200,1.5534053378634984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,65536,1.9482248094346788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,16384,0.4978213310241699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,12288,0.3868017726474338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,16384,0.8895609113905164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,10240,0.3108124468061659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,12288,0.6435457865397135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,8192,0.25042133861117893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,10240,0.5299004448784722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,65536,3.9524188571506076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,7168,0.21982755925920275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,8192,0.4204053348965115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,6144,0.18703644805484346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,51200,3.0889199574788413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,5120,0.15548443794250488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,7168,0.3668035666147868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,6144,0.31202933523390025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,4096,0.12480710612403022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,3584,0.11378222041659886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,4096,0.20731022622850206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,5120,0.2592906686994764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,3072,0.09767378038830227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,2560,0.08176000250710382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,3584,0.18152266078525117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,2048,0.06665333112080891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,3072,0.15944711367289224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,2560,0.13693422741360134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,2048,0.11312445004781087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,1024,0.03696799940533108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,768,0.029048889875411987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,1536,0.09053688579135472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,1024,0.06897155443827312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,512,0.02239911092652215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,768,0.05591999822192722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,256,0.01792266633775499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,128,0.016179554992251925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,512,0.0466577774948544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,256,0.03862133291032579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,64,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,32,0.01347733371787601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,128,0.03586311141649882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,64,0.035439110464519925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,12288,32,0.034619556532965765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,12288,1536,0.052114665508270264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,51200,1.2526417838202584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,65536,1.5526577631632488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,16384,0.4243155585394965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,12288,0.3137084378136529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,16384,0.7801600032382541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,10240,0.25585956043667263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,12288,0.5653840170966254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,51200,2.7169422573513455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,10240,0.47171820534600156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,7168,0.18188533518049452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,65536,3.474689695570204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,8192,0.3745608859592014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,6144,0.1695528825124105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,5120,0.1386622190475464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,7168,0.32659734619988334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,6144,0.27823111746046275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,4096,0.10960711373223199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,5120,0.22894755999247232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,8192,0.20683911111619738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,3584,0.09698133336173163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,3072,0.08332266410191853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,4096,0.18445955382453072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,3584,0.1621840000152588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,2560,0.07044977611965604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,3072,0.1420631143781874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,2048,0.057687997817993164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,1536,0.04435111085573832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,2560,0.1220853328704834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,1024,0.031247112486097548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,2048,0.10149777597851223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,768,0.02477866742346022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,1536,0.08142044146855672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,1024,0.06145422326193916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,512,0.019973332683245342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,256,0.016191111670600045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,768,0.04905688762664795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,128,0.013238222234778933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,256,0.033047109842300415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,64,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,128,0.03094489044613308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,10240,32,0.011527111132939657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,64,0.03105777833196852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,32,0.03046311272515191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,10240,512,0.04036710990799798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,65536,1.5105875862969294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,51200,1.1657181845770943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,16384,0.38937777943081325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,65536,2.9874729580349393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,16384,0.676838239034017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,12288,0.2971404393513997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,10240,0.230950223075019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,12288,0.4950373437669542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,8192,0.18630311224195692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,51200,2.3330525292290583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,7168,0.1646560033162435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,10240,0.4155928823682997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,6144,0.14197066095140246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,8192,0.3279244369930691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,5120,0.11622488498687744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,7168,0.2864017751481798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,4096,0.09603466590245564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,6144,0.2429386774698893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,3584,0.08444977468914455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,5120,0.20022043916914198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,3072,0.07358933157391019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,4096,0.16119466887580022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,2560,0.06169955597983467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,3584,0.14317244953579372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,2048,0.051190220647388034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,3072,0.12503733899858263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,1536,0.040595554643207125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,1024,0.026866666144794885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,2560,0.10815822415881687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,2048,0.08955199850930108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,768,0.020344000723626878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,1536,0.07110044691297743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,512,0.016207110550668504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,256,0.01256533298227522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,1024,0.05272266599867079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,768,0.042711112234327525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,128,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,512,0.03496088915401035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,256,0.028897778855429754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,128,0.02712000078625149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,32,0.010433778166770935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,64,0.027132444911532935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,8192,32,0.02681244413057963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,8192,64,0.00998133338159985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,51200,0.8545377519395617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,65536,1.1530799865722656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,16384,0.30826576550801593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,12288,0.2309511237674289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,16384,0.6253591113620335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,10240,0.19625955157809785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,12288,0.5059048864576552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,8192,0.1690408918592665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,10240,0.3830533292558458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,65536,2.71706051296658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,7168,0.13934132787916395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,8192,0.3045626746283637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,6144,0.12046399381425645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,7168,0.2654000123341878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,51200,2.1584915584988065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,5120,0.09909511274761623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,6144,0.2251502143012153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,4096,0.08371377653545803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,3584,0.07136533657709758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,5120,0.18564444118075898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,3072,0.06182399723264906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,4096,0.14936178260379368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,2560,0.05377689003944397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,3584,0.1315431065029568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,2048,0.0444435543484158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,1536,0.03438222077157762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,3072,0.11637866497039795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,2560,0.10085066821840073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,2048,0.08302577998903063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,1024,0.02436977790461646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,768,0.0200284438000785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,1536,0.06640711095598009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,512,0.015487111277050443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,1024,0.04902133345603943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,256,0.012349333200189801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,768,0.04018222292264303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,512,0.032709333631727434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,128,0.010408889088365767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,256,0.026510222090615168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,64,0.009085333181752099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,7168,32,0.009401777552233802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,128,0.025403555896547105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,64,0.025088000628683303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,7168,32,0.02514666650030348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,65536,0.9707155227661133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,51200,0.7755084567599826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,16384,0.26619288656446666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,16384,0.5872791078355577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,12288,0.1915795538160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,65536,2.507692548963759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,10240,0.16957688331604004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,12288,0.4249360031551785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,51200,1.9551484849717882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,8192,0.12939377625783285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,7168,0.11371733082665338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,10240,0.35403021176656085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,8192,0.28156089782714844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,6144,0.10217422246932983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,5120,0.08540978034337361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,7168,0.2459191216362847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,6144,0.2076328860388862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,5120,0.17100711663564047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,3584,0.061153776115841336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,4096,0.13750933276282415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,3072,0.052758223480648465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,3584,0.12167199452718098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,2560,0.04553600152333578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,3072,0.10620533095465766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,2048,0.03817866577042474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,2560,0.0927128858036465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,1536,0.03153511219554477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,2048,0.07686133517159356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,1024,0.023331556055280898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,4096,0.06838755475150214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,768,0.017661333084106445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,1536,0.05998311440149943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,512,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,1024,0.04493688874774509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,256,0.012482666307025485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,768,0.03744711147414313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,512,0.02979288829697503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,128,0.009413333402739631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,256,0.02473066747188568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,64,0.008035555481910706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,6144,32,0.008303999900817871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,128,0.023229334089491103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,64,0.023377777801619634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,6144,32,0.02313599983851115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,65536,0.8803697692023383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,51200,0.6576888826158311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,16384,0.23815290133158365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,12288,0.18117245038350424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,16384,0.5221306482950846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,10240,0.13772711488935682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,12288,0.406497769885593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,8192,0.11713955137464736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,10240,0.32584532101949054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,51200,1.77729066212972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,65536,2.2461200290256076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,7168,0.09931021928787231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,8192,0.2594044473436144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,6144,0.0832808878686693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,5120,0.0716480016708374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,4096,0.05747466617160373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,7168,0.22617332140604654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,5120,0.15702133708530003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,3584,0.05139288968510098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,6144,0.19149955113728842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,3072,0.045773333973354764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,2560,0.03803911142879062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,3584,0.11222933398352729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,2048,0.03091199861632453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,3072,0.09823555416531032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,1536,0.0251528885629442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,2560,0.08277955320146349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,2048,0.07048800256517199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,1024,0.018559111489189994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,768,0.014829332629839579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,1536,0.055360890097088285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,512,0.011522666447692446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,1024,0.041725334193971425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,256,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,768,0.034127109580569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,128,0.007996444072988298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,512,0.028043554888831243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,256,0.022950222094853718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,64,0.007290666302045186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,128,0.021394666698243883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,4096,0.12657422489590114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,5120,32,0.007356444166766272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,64,0.02142933342191908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,5120,32,0.020953777763578627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,65536,0.6503289010789659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,51200,0.5115911165873209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,16384,0.17403644985622832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,16384,0.4941946665445964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,12288,0.12738577524820963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,12288,0.35451467831929523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,65536,2.006835513644748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,51200,1.5806283950805664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,10240,0.11307999822828506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,8192,0.09317777554194133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,7168,0.08787288930681016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,10240,0.29680800437927246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,8192,0.2361839877234565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,6144,0.06946844524807401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,7168,0.20593778292338052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,5120,0.059083554479810924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,4096,0.04783822099367777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,5120,0.14275199837154812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,3584,0.04239733351601494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,4096,0.11608978112538655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,3072,0.03642844491534763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,3584,0.10261688629786174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,2560,0.03141422073046366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,3072,0.08993155426449245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,2048,0.025682666235499915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,2560,0.07564089033338758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,1536,0.02100533412562476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,6144,0.1744497749540541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,2048,0.0614559981558058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,1024,0.01612977849112617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,768,0.013607111242082385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,1536,0.05035110976960924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,512,0.010784000158309937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,1024,0.038776887787712946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,768,0.03067200051413642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,256,0.00867911097076204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,512,0.025096888343493145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,256,0.020612445142534044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,64,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,128,0.01960177719593048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,32,0.006665777828958299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,64,0.019322666856977675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,4096,32,0.019222221440739103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,65536,0.6372444364759657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,4096,128,0.0069217777086628815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,51200,0.5163715680440267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,16384,0.1717146635055542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,12288,0.12848622269100612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,16384,0.46472353405422634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,65536,1.9060550265842016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,10240,0.11090755462646484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,8192,0.08948089016808404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,12288,0.3376951217651367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,51200,1.4942826165093317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,7168,0.07795733213424683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,10240,0.28256532880995006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,6144,0.06610577636294894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,5120,0.054867555697758995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,8192,0.22489778200785318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,4096,0.04520888792143928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,7168,0.19600978162553576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,6144,0.1654426654179891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,3584,0.04140355520778232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,5120,0.1363555590311686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,3072,0.03506755497720506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,4096,0.10970133543014526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,2560,0.03074755602412754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,3584,0.09754400120841132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,3072,0.08550933334562515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,1536,0.021077333225144282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,2560,0.07197244299782647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,1024,0.014549333188268872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,2048,0.05808711051940918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,768,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,1536,0.0476817786693573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,1024,0.03607111175855001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,512,0.009307555854320526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,256,0.006626666833957036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,768,0.029193778832753498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,128,0.006687110910813014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,512,0.024248000648286607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,256,0.0199635558658176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,64,0.006265777680608962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,32,0.006631999793979857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,128,0.018593778212865193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,64,0.018544000056054857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3584,32,0.01829333272245195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3584,2048,0.025815111067559984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,65536,0.49214755164252383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,51200,0.38578934139675564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,16384,0.12880888250139025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,12288,0.10059022241168553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,16384,0.4415102269914415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,10240,0.08559288581212361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,12288,0.3213013278113471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,10240,0.2680853207906087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,51200,1.3903004328409831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,8192,0.06639377938376533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,65536,1.7569910685221355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,7168,0.05987111065122816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,6144,0.05158399873309665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,7168,0.18599466482798258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,8192,0.21419200632307264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,5120,0.043303999635908336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,6144,0.1574595504336887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,4096,0.03759288787841797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,3584,0.032518221272362605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,5120,0.12963022126091853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,4096,0.10374755329555935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,3072,0.02887466549873352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,3584,0.09212444225947063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,2560,0.025256888733969793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,2048,0.021106666988796655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,3072,0.07976000176535712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,1536,0.017511111166742113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,2560,0.06733689043256971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,1024,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,2048,0.055514666769239635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,768,0.011033777561452655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,1024,0.033599111768934466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,512,0.008661333057615492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,768,0.027983999914593164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,256,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,512,0.02312533391846551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,128,0.005953777581453323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,256,0.018957333432303537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,128,0.017658667431937326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,64,0.005305777821275923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,3072,32,0.005621333503060871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,64,0.01757688820362091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,32,0.017655111021465726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,3072,1536,0.04345866706636217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,65536,0.41732533772786456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,51200,0.33227290047539604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,16384,0.11523377895355225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,16384,0.41009243329366046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,12288,0.08524888753890991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,51200,1.2632897694905598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,65536,1.612132496303982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,12288,0.30382577578226727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,10240,0.07931733131408691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,8192,0.05928799841139051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,7168,0.050960001018312245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,8192,0.2028613355424669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,6144,0.04322310951020983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,7168,0.17535555362701416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,5120,0.037680887513690524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,6144,0.1482399966981676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,4096,0.03072355522049798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,5120,0.12224266264173721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,3584,0.02843733297453986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,4096,0.09898577796088324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,3584,0.08662933111190796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,3072,0.07521866427527533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,10240,0.2539973258972168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,2560,0.02115733259254032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,2048,0.017837334010336135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,1536,0.014768888552983602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,2560,0.06515022118886311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,2048,0.05272266599867079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,1024,0.010748444332016839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,1536,0.04115466607941522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,768,0.008807111117574904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,512,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,1024,0.031327111853493586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,256,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,768,0.025986666480700176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,128,0.005611555443869696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,512,0.02202933364444309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,64,0.004895110925038655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,3072,0.02441155579355028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2560,32,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,128,0.016488000750541687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,64,0.01663466625743442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,32,0.01670488880740272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,65536,0.3263733386993408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,51200,0.2531368997361925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,16384,0.08666400114695232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,16384,0.37901867760552305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2560,256,0.017850667238235474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,12288,0.06518399715423584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,10240,0.05836177534527249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,51200,1.1737360424465604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,12288,0.2868444389767117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,8192,0.04830933279461331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,10240,0.23995465702480742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,7168,0.04227111074659559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,65536,1.503409809536404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,6144,0.03593066665861342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,5120,0.029807110627492268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,8192,0.19195732805464003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,4096,0.02451466686195797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,7168,0.16655200057559544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,3584,0.02169422143035465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,6144,0.14059378041161433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,5120,0.11535467041863336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,3072,0.019329778022236295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,4096,0.0931439995765686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,2560,0.017090666625234816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,3584,0.08317333459854126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,2048,0.014485332700941296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,3072,0.0719733304447598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,1536,0.011782222323947482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,2560,0.06085422303941515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,1024,0.009156444834338294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,2048,0.0498924454053243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,1536,0.03868622250027127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,768,0.008361777497662438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,1024,0.029303110308117334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,512,0.00700444479783376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,256,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,768,0.024708444873491924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,512,0.02035911050107744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,128,0.0052044445441828836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,64,0.004658666749795278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,256,0.01700888905260298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,2048,32,0.004621333132187526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,128,0.015589333242840238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,64,0.015542222393883599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,2048,32,0.015543111496501498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,65536,0.2558151086171468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,51200,0.20910488234625924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,16384,0.06770310799280803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,16384,0.3591920004950629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,12288,0.05486577749252319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,51200,1.0950497521294487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,65536,1.3977866702609594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,10240,0.04681333237224155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,8192,0.037419554260041975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,12288,0.2710435655381945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,7168,0.032744000355402626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,10240,0.22676889101664224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,8192,0.18084088961283365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,6144,0.02769244379467434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,5120,0.023451555106374953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,6144,0.13241067197587755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,7168,0.15704711278279623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,4096,0.019923556182119582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,5120,0.10817866855197483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,3584,0.017976888351970248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,3072,0.016502221425374348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,4096,0.0881893303659227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,3584,0.07773600021998088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,2560,0.01463733282354143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,3072,0.06799644231796265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,2048,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,1536,0.010416000253624385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,2560,0.057152887185414634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,1024,0.00793333351612091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,2048,0.04607999987072415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,768,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,1536,0.03713422351413303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,512,0.00600000015563435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,1024,0.02793866727087233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,256,0.004981333182917701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,768,0.023295111126369897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,128,0.004550222307443619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,512,0.019918221566412184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,256,0.01588711142539978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,64,0.004307555655638377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,128,0.014470222923490735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1536,32,0.004287999951177173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,64,0.014519999424616495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1536,32,0.014147554834683737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,65536,0.17652266555362275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,51200,0.1389626661936442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,16384,0.04981155527962578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,16384,0.33697599834865993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,12288,0.03600177831119961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,51200,1.0255147086249459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,12288,0.25528889232211643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,10240,0.030722667773564655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,65536,1.3072018093532987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,8192,0.025014221668243408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,7168,0.02256533337963952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,10240,0.21374932924906412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,8192,0.17046222421858045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,6144,0.01962844365172916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,7168,0.14778399467468262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,4096,0.014367111855083041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,6144,0.12438489331139459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,5120,0.10205511252085368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,3584,0.013272888958454132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,3072,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,4096,0.08281333578957452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,2560,0.010127999716334874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,3584,0.07295644283294678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,2048,0.008922666311264038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,3072,0.06351644463009305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,1536,0.007984889050324758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,2560,0.05339733428425259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,2048,0.04256800148222181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,1024,0.006274666637182236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,768,0.005560889012283749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,1536,0.03416000141037835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,1024,0.02610933283964793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,512,0.005000000198682149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,768,0.02200266718864441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,5120,0.01679733395576477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,256,0.004586666822433472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,128,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,512,0.018568888306617737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,64,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,128,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,256,0.015015999476114908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,1024,32,0.003740444365474913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,64,0.01349244432316886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,65536,0.14805777867635092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,1024,32,0.013606222139464485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,51200,0.11802044179704453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,16384,0.0409804450141059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,16384,0.32682845327589244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,12288,0.03035022152794732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,51200,0.9911653730604383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,12288,0.24742489390903047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,10240,0.026163554853863184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,8192,0.021401777863502502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,10240,0.2078026665581597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,7168,0.019313777486483257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,7168,0.142920003996955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,6144,0.017242666747834947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,6144,0.12074755297766791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,5120,0.014549333188268872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,65536,1.2646222644382052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,4096,0.012119111087587146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,5120,0.09844355450736152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,3584,0.011601777540312873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,4096,0.0789288878440857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,3072,0.01000711073478063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,8192,0.1655164427227444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,2560,0.008630221916569604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,3584,0.07049777772691515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,2048,0.007961778177155389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,3072,0.06103110975689358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,1536,0.0069439998931354946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,2560,0.05070844292640686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,1024,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,2048,0.04116800096299913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,768,0.005219555563396878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,1536,0.03282577792803446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,1024,0.025431111454963684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,512,0.004543111142185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,256,0.0041751112374994485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,768,0.021387555532985266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,128,0.0035777779089079965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,512,0.01827555563714769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,256,0.014164444473054675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,64,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,768,32,0.003629333443111844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,128,0.013129777378506131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,64,0.0127368892232577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,768,32,0.013087111214796701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,65536,0.14388533433278403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,51200,0.11376444498697917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,16384,0.03966222206751505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,16384,0.31714044676886666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,12288,0.029973334736294214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,51200,0.9597244262695312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,10240,0.02294755478700002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,12288,0.24015378952026367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,65536,1.2234515084160698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,8192,0.019149333238601685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,7168,0.01705333259370592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,10240,0.20140443907843697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,8192,0.160243550936381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,6144,0.015057777365048727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,7168,0.13908178276485866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,5120,0.013045333325862885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,4096,0.011018666956159802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,6144,0.11698667208353679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,5120,0.0956968863805135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,3584,0.009749333063761393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,4096,0.07669511106279162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,3072,0.008984888593355814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,2560,0.007976888782448238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,3584,0.06806221935484144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,2048,0.007268444531493717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,3072,0.05879288911819458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,1536,0.006353777729802662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,2048,0.039431111680136785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,1536,0.03161244591077169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,1024,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,768,0.004879999905824661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,512,0.004287999951177173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,1024,0.02458133300145467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,256,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,512,0.01791555517249637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,128,0.00351555562681622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,256,0.014021333720948962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,64,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,128,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,64,0.012427555190192329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,512,32,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,65536,0.07464622126685248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,2560,0.04881422387229072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,32,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,51200,0.06040800280041165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,512,768,0.020943111843532983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,16384,0.023430221610599097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,16384,0.30601512061225045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,12288,0.019388443893856473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,51200,0.9298906326293945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,65536,1.1855991151597765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,10240,0.017708443933063082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,8192,0.014528888795110913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,12288,0.2326400015089247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,7168,0.013117333253224691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,10240,0.1946480009290907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,8192,0.15597244103749594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,6144,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,5120,0.01033688916100396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,7168,0.13483999835120306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,4096,0.008013333711359236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,6144,0.11316000090705024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,5120,0.0925368865331014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,4096,0.0743137796719869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,3072,0.0069164443347189165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,3584,0.06527911292182074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,2560,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,3072,0.05643288956748115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,2048,0.005889777921968036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,2560,0.046688887808057994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,1536,0.004885333279768626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,2048,0.037520001331965126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,1024,0.004536889079544279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,1536,0.030304001437293157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,768,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,1024,0.023919110496838886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,3584,0.007996444072988298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,512,0.003839111162556542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,768,0.020259555843141343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,256,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,128,0.0031644445326593188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,512,0.016905777984195285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,256,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,64,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,256,32,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,128,0.01238044434123569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,65536,0.06768533256318834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,32,0.011809777882364062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,51200,0.05301510956552294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,51200,0.9007750617133247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,65536,1.1484764946831596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,256,64,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,16384,0.013883555928866068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,12288,0.013467555244763693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,10240,0.012412444584899478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,12288,0.2255208757188585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,16384,0.29743554857042104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,8192,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,10240,0.18928089406755236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,7168,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,6144,0.010794666906197866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,8192,0.1514471107059055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,5120,0.009312888814343346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,7168,0.13078667057885066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,6144,0.10957955651813084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,5120,0.08908533387713963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,3584,0.007685333490371704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,4096,0.07171911001205444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,3072,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,3584,0.06333777639600965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,2560,0.005879111174080107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,3072,0.054225779241985746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,2048,0.005264889034960005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,2560,0.04521244433191088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,1536,0.004874666531880696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,2048,0.03753688931465149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,1024,0.004219555606444676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,1536,0.029827555020650227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,768,0.0038497779104444715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,1024,0.02332088847955068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,512,0.00350222239891688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,768,0.019554666346973844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,512,0.016564443707466125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,128,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,256,0.013079110946920184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,64,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,128,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,32,0.00300444434914324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,64,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,65536,0.05981510877609253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,128,32,0.011377777490350934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,51200,0.04723466767205132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,128,4096,0.008619555168681675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,16384,0.01239466667175293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,16384,0.2968382305569119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,12288,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,51200,0.900218645731608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,10240,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,65536,1.1480258305867512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,8192,0.010312000082598792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,12288,0.2254951000213623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,10240,0.18932089540693495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,7168,0.009759999811649323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,8192,0.1513626707924737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,6144,0.009700444837411245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,7168,0.1303608947330051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,6144,0.1086568898624844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,5120,0.00868977771864997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,4096,0.007356444166766272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,5120,0.0887404415342543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,3584,0.0069288888739215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,4096,0.07158311208089192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,3072,0.006254222244024277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,3584,0.06337866518232557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,2560,0.005884444548024072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,3072,0.0539902216858334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,2048,0.005223111146026188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,2560,0.04520888792143928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,1536,0.004824888788991504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,2048,0.03644533289803399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,1024,0.004193777839342753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,1536,0.02968888812594944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,768,0.0038533334930737815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,1024,0.02307111190425025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,512,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,768,0.019553777244355943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,256,0.0031600000543726813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,512,0.016513778103722464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,128,0.003142222141226133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,256,0.012913777596420713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,64,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,128,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,64,32,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,64,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,64,32,0.011388444238238864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,65536,0.058155556519826256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,51200,0.045540445380740695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,51200,0.8997910817464193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,16384,0.011694221860832639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,65536,1.148249732123481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,12288,0.010008888940016428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,10240,0.009330666727489894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,12288,0.22556710243225098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,10240,0.18909154997931588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,8192,0.008634666601816813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,7168,0.008209777375062307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,16384,0.29723021719190806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,8192,0.15127289295196533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,6144,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,7168,0.1302622159322103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,5120,0.010399999717871347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,4096,0.009034666750166152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,6144,0.10881333218680488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,5120,0.08895199828677708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,3584,0.008347555167145198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,3072,0.007619555625650618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,4096,0.07227555248472425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,2560,0.006685333533419504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,3584,0.0633680025736491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,3072,0.05405422382884555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,2048,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,1536,0.004604444321658877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,2560,0.044970668024486966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,2048,0.0358168880144755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,1024,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,1536,0.0296382208665212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,768,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,512,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,1024,0.02300888962215847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,256,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,768,0.019661333825853135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,512,0.016546666622161865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,64,0.0029360000044107437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,256,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2561,32,32,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,128,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,64,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2561,32,32,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,16384,2.6235129038492837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,16384,3.624988555908203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,51200,7.828704833984375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,12288,1.9990098741319444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,16384,2.701829274495443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,12288,2.5253751542833114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,51200,12.634769015842013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,51200,8.490393744574652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,10240,1.6308808856540258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,12288,2.022650612725152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,10240,2.075162675645616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,8192,1.2694418165418837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,8192,1.6157315572102864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,10240,1.71052000257704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,7168,1.1100133260091145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,8192,1.354355600145128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,7168,1.3925662570529516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,6144,0.9292568630642362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,6144,1.1838133070203993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,7168,1.1843333774142795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,5120,0.7803484598795573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,4096,0.6200470924377441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,6144,1.0232808854844835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,5120,0.986146714952257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,5120,0.8586924340989855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,4096,0.7798239919874402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,3584,0.5212071206834581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,4096,0.702547550201416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,3584,0.6842942237854004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,3072,0.4876711103651259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,3584,0.6220693588256836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,3072,0.5918915536668565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,2560,0.3981822331746419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,3072,0.5480613178677028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,2560,0.5046613481309679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,2048,0.3323422273000081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,2560,0.4655111100938585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,2048,0.41237955623202854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,1536,0.24377332793341744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,2048,0.3907039960225423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,1536,0.3262115584479438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,1024,0.1692213349872165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,1536,0.3118142286936442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,1024,0.24758399857415092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,768,0.1340586609310574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,1024,0.24789865811665854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,768,0.19843822055392793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,768,0.19695466094546846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,512,0.10006044308344524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,256,0.07535022497177124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,512,0.16219377517700195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,512,0.16876088248358834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,128,0.06568977567884657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,256,0.14945067299736872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,256,0.12773066096835667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,64,0.05533777674039205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,128,0.14169332716200086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,65536,128,0.10148621930016412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,65536,32,0.0573271115620931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,64,0.14340266916486952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,65536,32,0.1448648903104994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,51200,6.096010843912761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,65536,7.977440728081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,65536,12.728697035047743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,65536,8.89265611436632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,16384,2.0374107360839844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,16384,2.9117084079318576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,51200,10.016120062934027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,16384,2.1906471252441406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,51200,6.855687035454644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,12288,1.625485314263238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,12288,2.0111573537190757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,10240,1.3089635637071397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,10240,1.653132438659668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,12288,1.6288515726725261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,8192,1.0241422653198242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,10240,1.3613671196831598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,7168,0.8686453501383463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,8192,1.2787502076890733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,8192,1.0997716055976021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,6144,0.7779733339945475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,7168,1.115655157301161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,5120,0.5983697573343912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,7168,0.9673902723524305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,6144,0.942930645412869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,4096,0.5175182024637858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,6144,0.8321866459316678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,5120,0.7893679936726888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,5120,0.7005439864264594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,4096,0.6454319953918457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,3584,0.4218124548594157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,4096,0.5744249025980631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,3072,0.39503200848897296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,3584,0.5496648682488335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,3584,0.5099822150336372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,2560,0.31405332353379994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,3072,0.4766986634996202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,3072,0.4467048645019531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,2048,0.25464267200893825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,2560,0.40526578161451554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,2560,0.38022221459282773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,1536,0.18779377142588297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,2048,0.33217954635620117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,1024,0.13424355453915066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,1536,0.2623644404941135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,2048,0.31866667005750865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,1536,0.25647465387980145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,768,0.1163440015580919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,1024,0.19953155517578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,1024,0.19639644357893202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,512,0.07885866694980197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,768,0.15859911176893446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,768,0.16381155120001897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,256,0.061195552349090576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,512,0.13593067063225642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,512,0.1344453361299303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,128,0.05099022057321337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,256,0.10591466559304132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,256,0.1195911169052124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,64,0.044340444935692676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,51200,128,0.08394755257500543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,128,0.113590227233039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,51200,32,0.046535111135906644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,64,0.11572978231641982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,51200,32,0.11650044388241237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,65536,2.619262271457248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,51200,2.0362870958116317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,65536,4.785511016845703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,65536,3.814849005805122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,16384,0.6836622026231555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,16384,1.078639136420356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,12288,0.5023440255059136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,16384,0.9710711373223199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,51200,3.7738126118977866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,10240,0.42049598693847656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,12288,0.7741582128736707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,51200,2.9892124599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,12288,0.7365626758999295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,10240,0.667717350853814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,8192,0.34321954515245223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,10240,0.6183084381951226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,8192,0.504768901401096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,7168,0.29665689998202854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,6144,0.2478275563981798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,8192,0.5013555420769585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,7168,0.43995200263129336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,7168,0.4387839900122748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,5120,0.19565155771043566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,6144,0.3766728772057428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,6144,0.3736204571194119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,5120,0.3099386692047119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,4096,0.17139377858903673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,5120,0.31724977493286133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,3584,0.14774755636850992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,4096,0.24830312199062773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,4096,0.2600453429751926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,3072,0.12579822540283203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,3584,0.21826933489905465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,2560,0.10539911190668742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,3584,0.23241066932678223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,3072,0.19037866592407227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,3072,0.20433333184983996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,2048,0.08624177508884007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,2560,0.16339288817511663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,2560,0.18503022193908691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,1536,0.06544711192448933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,2048,0.1354853312174479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,2048,0.14747733540005153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,1536,0.1074977782037523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,1024,0.046330667204327054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,1536,0.11976710955301921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,768,0.0370142228073544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,1024,0.09366222222646077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,1024,0.08232622014151679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,512,0.029265777932273015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,768,0.06638844145668878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,768,0.07919733391867743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,256,0.021777777208222285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,512,0.06650577651129828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,128,0.019889778561062284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,256,0.05285066697332594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,256,0.0490497781170739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,64,0.0229111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,16384,128,0.04408355553944906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,128,0.04550933175616794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,16384,32,0.023689778314696416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,32,0.04626577761438158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,64,0.046718223227394946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,16384,512,0.05655644337336222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,65536,2.0821270412868924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,51200,1.6840044657389324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,65536,3.8815633985731335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,16384,0.5498711268107096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,65536,3.291524463229709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,51200,3.035054312811958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,16384,0.8635013368394641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,12288,0.4019324514600966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,16384,0.841932455698649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,51200,2.581603580050998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,12288,0.6296719974941677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,10240,0.3314266734653049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,12288,0.6367857721116807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,8192,0.2692506578233507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,10240,0.5222800042894151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,7168,0.229967991511027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,8192,0.4126613405015733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,8192,0.4325848950280084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,7168,0.36016178131103516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,6144,0.19719910621643066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,7168,0.3779466682010227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,5120,0.1600666708416409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,6144,0.3068942228953044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,6144,0.325180451075236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,4096,0.1307351059383816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,5120,0.2538373205396864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,5120,0.27383555306328666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,10240,0.5360035366482204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,3584,0.11499288347032334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,4096,0.2230151229434543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,4096,0.20376000139448378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,3584,0.17928889062669542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,3072,0.09982222318649292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,3584,0.19993333021799722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,2560,0.08452177710003322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,3072,0.17557866043514678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,3072,0.15720266766018337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,2048,0.06615555286407471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,2560,0.13511021931966147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,2560,0.15150754981570774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,1536,0.055310222837660045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,2048,0.11207021607293023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,2048,0.1276924477683173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,1024,0.037943111525641546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,1536,0.08913511037826538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,1536,0.10468710793389215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,1024,0.06827288866043091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,768,0.030654221773147583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,1024,0.08142133553822835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,512,0.02478666603565216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,768,0.05460355679194132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,768,0.06926844517389934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,256,0.01852799952030182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,512,0.045597334702809654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,512,0.05920444594489204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,128,0.015545777148670621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,256,0.04643644558058845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,256,0.03850222296184964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,64,0.019353777170181274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,128,0.035664889547559954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,12288,128,0.038402666648228966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,12288,32,0.020590222544140287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,32,0.033971554703182645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,12288,64,0.034648001194000244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,65536,1.6914212968614366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,51200,1.2795546849568684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,65536,3.4069493611653647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,65536,3.0308755238850913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,16384,0.4464844597710504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,51200,2.6633546617296004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,16384,0.7673173480563693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,16384,0.7749030854966906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,12288,0.32325866487291127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,10240,0.2603030999501546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,12288,0.5586666531032985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,51200,2.3798319498697915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,12288,0.5876302189297146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,10240,0.46357864803738064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,8192,0.2347093423207601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,10240,0.4952809015909831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,7168,0.20207556088765463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,8192,0.3673333326975505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,6144,0.1653591129514906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,8192,0.401184876759847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,7168,0.3219875494639079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,7168,0.36220622062683105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,5120,0.13756088415781656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,6144,0.2724382347530789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,6144,0.30005865626864964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,4096,0.10971733596589829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,5120,0.2246639993455675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,5120,0.2519893381330702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,3584,0.09925955533981323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,4096,0.20587733056810167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,4096,0.18091377947065565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,3072,0.08563733100891113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,3584,0.15999733077155218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,3584,0.18374755647447374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,2560,0.07001777489980061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,3072,0.14025155703226724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,3072,0.16197333070966932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,2048,0.05784444676505195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,2560,0.12079289224412705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,1536,0.04425244530042013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,2560,0.1406515571806166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,2048,0.09992711411582099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,2048,0.11810933219061957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,1024,0.031082666582531396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,1536,0.07975644535488553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,1536,0.0970275534523858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,768,0.024812445044517517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,1024,0.06037777662277222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,1024,0.07592711183759901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,512,0.01990133358372582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,768,0.06412355767356025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,768,0.04868888854980469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,512,0.03936977850066291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,256,0.018067555295096505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,512,0.05431644452942742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,128,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,256,0.04382755690150791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,10240,128,0.03403288788265652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,128,0.03070577647950914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,32,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,64,0.030277334981494482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,32,0.030048890246285334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,10240,64,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,65536,1.2539884779188368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,10240,256,0.03265600072013007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,51200,0.9937956068250867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,65536,2.9328320821126304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,16384,0.3435484568277995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,16384,0.6818559964497884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,65536,2.7700826856825085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,51200,2.174861272176107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,12288,0.2695937686496311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,16384,0.7106177541944715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,51200,2.285708533393012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,12288,0.48894664976331925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,10240,0.23166577021280924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,12288,0.538332462310791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,8192,0.18766489293840197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,10240,0.40537601047092015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,10240,0.4539493454827203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,8192,0.32306843333774143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,7168,0.16324978404574925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,8192,0.36658843358357746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,7168,0.28211821450127494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,6144,0.14050311512417263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,7168,0.31915556059943306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,5120,0.1155431138144599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,6144,0.2387697696685791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,6144,0.27371467484368217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,4096,0.0938462217648824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,5120,0.2299591170416938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,5120,0.19740711318122017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,4096,0.15786043802897134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,3584,0.08075022035174899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,4096,0.18837866518232557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,3072,0.07124800152248807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,3584,0.1407795614666409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,3584,0.16885777314503989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,2560,0.06084799766540527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,3072,0.12362221876780193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,3072,0.14859110779232448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,2048,0.050224890311559044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,2560,0.106532441245185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,2048,0.08831822209888035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,1536,0.03789600067668491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,2048,0.1109057797325982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,1536,0.06995111041598849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,1024,0.02548355526394314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,1536,0.0894755588637458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,1024,0.05240089032385084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,768,0.020648888415760465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,2560,0.12823022736443415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,768,0.04159555501408047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,512,0.015948444604873657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,1024,0.07055555449591742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,768,0.05971466832690769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,256,0.013085333009560903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,512,0.03388533327314589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,512,0.051035556528303355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,256,0.038378665844599404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,256,0.0285537772708469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,128,0.01126844435930252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,64,0.009755555954244401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,8192,128,0.03189511100451151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,128,0.02685244381427765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,8192,32,0.009964444571071202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,32,0.026196445027987163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,8192,64,0.027141332626342773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,65536,1.1177004708184135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,51200,0.8739440176222059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,65536,2.675143135918511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,16384,0.29549066225687665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,51200,2.1158150566948786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,65536,2.6452293395996094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,16384,0.6134533352322048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,12288,0.2349217732747396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,51200,2.0683990054660373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,16384,0.6795244216918945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,12288,0.4529075622558594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,10240,0.19220533635881212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,8192,0.15844621923234728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,12288,0.5145795610215929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,10240,0.37814754909939235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,10240,0.43328534232245547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,7168,0.13347821765475804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,8192,0.3495306703779433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,8192,0.30776090092129177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,6144,0.11686489317152236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,7168,0.262152009540134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,7168,0.3044568962521023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,5120,0.09291733635796441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,6144,0.26087021827697754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,6144,0.22299554612901476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,4096,0.07880622148513794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,5120,0.18245955308278403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,3584,0.06826311349868774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,4096,0.179639114273919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,4096,0.14698133203718397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,3584,0.13023911582099065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,3072,0.05762222078111437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,3584,0.1601697736316257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,2560,0.050494223833084106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,3072,0.11530755625830756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,2560,0.09945778052012126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,3072,0.1422248946295844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,2048,0.03978133201599121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,2560,0.12264622582329644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,1536,0.031946665710873075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,5120,0.2194382217195299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,2048,0.10329155789481269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,2048,0.08304533031251696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,1536,0.06492177645365398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,1024,0.022429333792792425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,1536,0.08483733071221246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,768,0.019155555301242404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,1024,0.048370665974087186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,1024,0.06717689169777764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,768,0.03887644410133362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,512,0.013752000199423896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,768,0.057160000006357826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,256,0.011165333290894827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,512,0.03194133440653483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,512,0.048927999205059476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,256,0.026503110925356548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,256,0.03670933180385166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,128,0.010392000277837118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,64,0.008990222381220924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,128,0.0247662216424942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,7168,128,0.030534221066368952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,7168,32,0.009374221993817223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,32,0.0243857784403695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,7168,64,0.025176000263955858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,65536,1.045104874504937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,51200,0.8099244435628256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,65536,2.4670772552490234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,65536,2.5103306240505643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,16384,0.2807404465145535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,51200,1.9257325066460504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,16384,0.5632978015475804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,16384,0.646682686275906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,12288,0.2056737740834554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,51200,1.9684071011013453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,12288,0.41826666726006406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,10240,0.17865422036912706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,12288,0.4907689094543457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,8192,0.13639822271135119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,10240,0.34918755955166286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,7168,0.10881600115034316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,10240,0.43132000499301487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,8192,0.2787048816680908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,8192,0.33427998754713273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,7168,0.24236978424919975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,6144,0.0994666682349311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,7168,0.28906310929192436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,5120,0.08284711175494723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,6144,0.20556888315412733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,6144,0.24835199779934355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,4096,0.06626666916741265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,5120,0.16840977138943145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,5120,0.20885688728756377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,3584,0.059602664576636426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,4096,0.13573155138227674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,4096,0.17119022210439047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,3072,0.048623998959859215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,3584,0.15250933170318604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,2560,0.04398133357365926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,3584,0.12066488795810276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,3072,0.10568533341089885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,3072,0.13468710581461588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,2048,0.03456177645259433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,2560,0.09238844447665745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,2048,0.0764933360947503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,2560,0.11736622121598984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,2048,0.09880977869033813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,1024,0.020607110526826646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,1536,0.059917330741882324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,1024,0.04512088828616672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,1536,0.08139466577106051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,768,0.016275554895401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,1024,0.06399110953013103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,512,0.014186667071448432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,768,0.0548257761531406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,512,0.030177778667873804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,768,0.03677689035733541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,256,0.011696000066068439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,512,0.04579288760821024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,128,0.00904444439543618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,256,0.03530577818552653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,128,0.022954665952258643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,256,0.024711110525661047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,1536,0.026751110951105755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,64,0.011109333071443768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,6144,128,0.03017510970433553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,6144,32,0.011135111252466837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,32,0.02274666726589203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,6144,64,0.02310755517747667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,65536,0.8668266402350532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,51200,0.6841875712076823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,65536,2.2343485090467667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,51200,1.7377760145399306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,65536,2.3822417789035373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,16384,0.5150835778978136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,51200,1.8654727935791016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,12288,0.16470932960510254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,16384,0.23106755150689018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,10240,0.13680711057451037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,12288,0.38506576750013566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,12288,0.4649137920803494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,8192,0.11431377463870579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,10240,0.3205191029442681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,10240,0.3923039966159397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,8192,0.25570132997300893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,7168,0.09972000122070312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,16384,0.6132657792833116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,8192,0.3176097869873047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,6144,0.08574133449130589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,7168,0.22224799791971842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,7168,0.27583556705051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,5120,0.07351466682222155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,6144,0.23552799224853516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,6144,0.18884889284769693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,5120,0.15528355704413518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,4096,0.05695111221737332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,5120,0.19769244723849824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,3584,0.05271111263169182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,4096,0.125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,4096,0.16241955757141113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,3584,0.11081688933902317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,3072,0.044933332337273486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,3584,0.1447226603825887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,2560,0.038113776180479265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,3072,0.09700710905922784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,3072,0.12864532735612658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,2560,0.08283378018273248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,2048,0.03148533238304986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,2560,0.11046489079793294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,1536,0.0247697780529658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,2048,0.07028800249099731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,2048,0.09384800328148736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,1536,0.07707822322845459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,1024,0.018395556343926322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,1024,0.04127022292878892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,768,0.014613333675596448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,768,0.03366400135887994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,768,0.05216266711552938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,512,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,1536,0.05478577812512716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,512,0.027803556786643133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,256,0.010042666561073726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,512,0.042261332273483276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,256,0.022721777359644573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,1024,0.06130311224195692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,128,0.009163555171754625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,256,0.03412444392840067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,64,0.008391111261314815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,5120,128,0.029504888587527808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,5120,32,0.008381333616044786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,128,0.021394666698243883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,64,0.020988444487253826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,5120,32,0.020976000361972388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,65536,0.7644649081759982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,51200,0.5943546824985081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,65536,2.0096374087863498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,16384,0.2144693268669976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,51200,1.5548630820380316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,16384,0.4828231069776747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,12288,0.1629528866873847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,12288,0.3512524498833551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,51200,1.7630479600694444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,16384,0.580692450205485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,10240,0.12145155005984837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,10240,0.2930897871653239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,65536,2.2504560682508683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,8192,0.08984266387091742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,12288,0.4409297837151422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,7168,0.08380177948209976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,10240,0.3719920052422418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,8192,0.23464621437920463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,8192,0.2998044490814209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,6144,0.07093244128757052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,7168,0.20322044690450033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,7168,0.26163466771443683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,5120,0.058436446719699435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,6144,0.17166666189829508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,6144,0.22328445646497938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,4096,0.045844445625940956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,5120,0.1873662206861708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,3584,0.042103111743927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,5120,0.14181599352094862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,4096,0.11415021949344212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,3072,0.037089778317345515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,3584,0.10085066821840073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,3584,0.1369431151284112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,2560,0.030858665704727173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,3072,0.08840799993938869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,3072,0.12146843804253472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,2048,0.025638222694396973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,2560,0.10548888974719578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,2048,0.06021510892444187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,2560,0.07454488674799602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,1536,0.02085955606566535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,2048,0.08874933587180243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,1024,0.014888889259762235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,1536,0.07285422086715698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,1536,0.04948088857862684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,1024,0.0374017788304223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,4096,0.1530782249238756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,768,0.01256444470749961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,1024,0.057111998399098717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,512,0.009379555781682333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,768,0.04794933398564657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,768,0.030305776331159804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,256,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,512,0.04079733292261759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,512,0.025340444511837427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,256,0.02068355513943566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,128,0.0070017774899800616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,256,0.03224622209866842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,64,0.0063884444534778595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,128,0.019293333093325298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,4096,128,0.02757866680622101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,4096,32,0.006331555545330048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,64,0.01937333411640591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,4096,32,0.019301333361201815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,65536,0.5823946528964573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,51200,0.4561857647365994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,65536,1.8521164788140192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,51200,1.4658204184638128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,16384,0.15168978108300105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,65536,2.1216142442491317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,16384,0.4462381998697917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,12288,0.11068977912267049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,16384,0.5497093200683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,51200,1.6602747175428603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,12288,0.3348337809244792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,10240,0.1020488871468438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,12288,0.41803733507792157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,8192,0.08184444242053561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,10240,0.27989421950446236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,10240,0.3515591091579861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,7168,0.06887466377682157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,8192,0.28476444880167645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,8192,0.2230453226301405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,6144,0.06016533242331611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,7168,0.1929173337088691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,7168,0.24670399559868705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,5120,0.05013244350751241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,6144,0.16452711158328587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,6144,0.20999111069573295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,4096,0.03958844476275974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,5120,0.13515733347998724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,5120,0.17709955904218885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,4096,0.10874044232898289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,3584,0.036786665519078575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,4096,0.14525689019097224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,3072,0.03169599837727017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,3584,0.09644533528221978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,3584,0.129856891102261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,3072,0.08441688617070515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,2560,0.027294221851560805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,3072,0.11460622151692708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,2048,0.02256533337963952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,2560,0.07146666447321574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,1536,0.018704000446531508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,2560,0.1004106667306688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,2048,0.05692444245020548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,2048,0.08411644564734565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,1024,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,1536,0.047028445535235934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,768,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,1536,0.0692782203356425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,1024,0.05438844362894694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,512,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,768,0.028652442826165095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,768,0.04725066820780436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,512,0.02384000023206075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,256,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,512,0.03876266545719571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,128,0.00666222224632899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,256,0.01956622302532196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,256,0.031962666246626116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,64,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,128,0.018294221825069852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3584,32,0.005986666513813867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,32,0.018240888913472492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,64,0.01835555500454373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3584,1024,0.035169776942994856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,65536,0.5355733235677084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3584,128,0.027057778504159715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,51200,0.4386044343312581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,65536,1.732629352145725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,16384,0.14223110675811768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,51200,1.3732710944281683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,65536,2.118904961480035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,16384,0.42290398809644913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,12288,0.11144533422258164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,51200,1.6620328691270616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,16384,0.548384878370497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,10240,0.08731733428107367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,12288,0.3189013269212511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,12288,0.4167991214328342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,10240,0.26595910390218097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,8192,0.07763644721772936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,10240,0.35184089342753094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,8192,0.2119528849919637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,8192,0.2835804356469048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,7168,0.1849342187245687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,6144,0.058096885681152344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,7168,0.24565333790249297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,6144,0.1560302178064982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,6144,0.2102666695912679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,5120,0.04860799842410617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,5120,0.12846577167510986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,5120,0.17675111028883192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,4096,0.039537777503331505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,7168,0.06975466675228544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,3584,0.03618311219745212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,4096,0.14447111553615993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,4096,0.10370755857891506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,3072,0.03128444486194187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,3584,0.12919910748799643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,3584,0.09193688631057739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,3072,0.08023199770185682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,2560,0.027144889036814373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,3072,0.11471377478705512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,2048,0.022984888818528917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,2560,0.06785866949293348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,2560,0.09935822089513142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,2048,0.05579377545250786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,2048,0.08344444301393296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,1536,0.018579555882347953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,1024,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,1536,0.042824887567096286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,1536,0.06901333067152235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,768,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,1024,0.05269066823853386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,1024,0.03396177623007033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,768,0.027444443768925134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,512,0.00942666663063897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,768,0.04459733433193631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,256,0.007335111498832703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,512,0.023153778579499986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,512,0.03849688834614224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,128,0.006337777607970768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,256,0.018943111101786297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,256,0.030769778622521296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,64,0.006314666734801398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,3072,128,0.026156443688604567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,128,0.01763911048571269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,64,0.01724088854259915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,3072,32,0.01770755483044518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,65536,0.4407324526045058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,3072,32,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,51200,0.33204976717631024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,65536,1.6274088753594294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,65536,1.9909644656711156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,51200,1.2868293126424153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,16384,0.11662488513522679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,16384,0.40068888664245605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,12288,0.08491288953357273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,16384,0.5162026617262099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,10240,0.075345774491628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,51200,1.5616693496704102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,12288,0.30158665445115834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,12288,0.3924604521857367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,8192,0.05335466729270088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,10240,0.25266488393147785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,7168,0.04972799950175815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,10240,0.3315075503455268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,8192,0.2012666728761461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,6144,0.04272177815437317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,8192,0.2670568890041775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,7168,0.17560799916585287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,5120,0.035805334647496544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,7168,0.23141243722703722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,6144,0.19773333602481416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,4096,0.02940355406867133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,5120,0.16628266705407038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,4096,0.09842933548821343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,3584,0.025080000360806782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,4096,0.13668533166249594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,3584,0.08666933245129055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,3072,0.022279999322361414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,3584,0.12206489510006374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,3072,0.07702044645945232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,6144,0.14807732899983725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,2560,0.01921422282854716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,3072,0.10867466529210408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,5120,0.12198132938808864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,2048,0.017436444759368896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,2560,0.06535022126303779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,2560,0.09336888790130615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,1536,0.013407111167907715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,2048,0.05266666743490431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,1024,0.010692444112565784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,2048,0.07918310827679105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,1536,0.040456000301573015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,1536,0.0640266670121087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,768,0.008791999684439765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,1024,0.031452443864610456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,1024,0.05158488949139913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,512,0.007684444387753804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,768,0.025789333714379206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,768,0.04282933473587036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,256,0.006330666856633292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,512,0.03806666533152262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,256,0.017865777015686035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,128,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,256,0.030223998758527968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,512,0.022117333279715642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,64,0.005611555443869696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2560,128,0.025938666529125635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,128,0.01650044487582313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2560,32,0.005592000153329637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,32,0.016519111063745286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2560,64,0.016570667425791424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,65536,0.4122684531741672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,51200,0.32272622320387095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,65536,1.4982577429877386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,51200,1.1729164123535156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,16384,0.10837688710954453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,65536,1.9914756351047094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,16384,0.3777022096845839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,12288,0.0819466643863254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,16384,0.5165031221177843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,12288,0.286281771130032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,10240,0.06559999783833821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,51200,1.55841244591607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,10240,0.23967822392781576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,12288,0.39331910345289445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,8192,0.05366666780577766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,10240,0.3299209011925591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,7168,0.04284088810284933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,8192,0.19104977448781332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,8192,0.26760000652737087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,6144,0.03936977850066291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,7168,0.16534311241573757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,7168,0.23126843240525988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,6144,0.1397040022744073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,6144,0.1973102225197686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,5120,0.031176888280444678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,4096,0.02624711063173082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,5120,0.11514666345384385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,5120,0.16533777448866102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,3584,0.022740445203251306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,4096,0.09308799770143296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,4096,0.13530666298336452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,3584,0.08240889178382026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,3072,0.021682666407691106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,3584,0.12189422713385688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,3072,0.07191644774542914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,3072,0.10786488983366226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,2560,0.017447999782032438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,2048,0.014498665928840637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,2560,0.06096800168355306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,2560,0.09370666742324829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,1536,0.012520000338554382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,2048,0.050119111935297646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,2048,0.07941333452860515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,1024,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,1536,0.03872888949182298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,1536,0.06517422199249268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,1024,0.02980177932315403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,768,0.008360000120268928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,768,0.02444355520937178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,512,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,768,0.04253777861595154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,256,0.006136888845099344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,512,0.03625866770744324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,512,0.020703110429975722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,128,0.005250666704442766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,256,0.0291920006275177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,256,0.016932444439993966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,64,0.0052622221410274506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,128,0.015513777732849121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,2048,32,0.00498844434817632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,64,0.015542222393883599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,2048,32,0.015145777000321282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,1024,0.049508445792728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,65536,0.2846817705366346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,51200,0.22364089224073622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,2048,128,0.02511466708448198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,65536,1.3954221937391493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,16384,0.07886222336027357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,51200,1.099789301554362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,16384,0.3564346631368001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,65536,1.8610719045003254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,12288,0.05434755484263102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,16384,0.48415289984809023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,10240,0.047275556458367236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,12288,0.2705555491977268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,51200,1.4574826558430989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,8192,0.03952266772588094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,10240,0.3111368815104167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,10240,0.2267164389292399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,8192,0.18071556091308594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,7168,0.035635554128222995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,8192,0.2519955635070801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,7168,0.15584622489081487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,6144,0.030609776576360066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,7168,0.21713956197102866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,5120,0.026163554853863184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,6144,0.1319591071870592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,6144,0.1852977805667453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,12288,0.36820088492499453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,5120,0.15571288267771402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,5120,0.10841777589586045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,4096,0.08719111151165432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,3584,0.019701333509551156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,4096,0.1270728905995687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,3584,0.07766844166649713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,3072,0.0173333336909612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,3584,0.11422755983140732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,3072,0.0680826637479994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,3072,0.10109688838322957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,2560,0.015138667490747241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,2560,0.05661244524849785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,2560,0.08807644579145645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,2048,0.013068444199032254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,2048,0.04648088746600681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,2048,0.07378488779067993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,1536,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,1536,0.0369742214679718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,1536,0.05985866652594673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,1024,0.008003555238246918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,1024,0.02787733409139845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,1024,0.04710400104522705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,768,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,768,0.04012711180580987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,512,0.006239999913507038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,4096,0.022245332598686218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,512,0.019331556227472093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,256,0.0052062223354975385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,512,0.03496355480617947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,256,0.015506666567590503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,256,0.028070221344629925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,128,0.004578666554556953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,128,0.014483556151390076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1536,128,0.02437422176202138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,64,0.004530666602982415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1536,32,0.004326222257481681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,64,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,32,0.014175110393100314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1536,768,0.02332088847955068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,65536,0.19084178076850045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,51200,0.14341155687967935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,51200,1.0231599807739258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,65536,1.8593209584554036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,16384,0.05345777670542399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,16384,0.3359804418351915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,65536,1.3066559897528756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,12288,0.03950044512748718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,51200,1.4571208953857422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,12288,0.2552337646484375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,16384,0.48303646511501735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,10240,0.03307111064592997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,12288,0.36804623074001735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,8192,0.026339555780092876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,10240,0.21369600296020508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,10240,0.3105448881785075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,7168,0.023910222782029047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,8192,0.17008088694678414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,8192,0.2513084411621094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,6144,0.020656888683636982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,7168,0.1477831072277493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,5120,0.01854933301607768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,7168,0.2168311145570543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,6144,0.12436266740163167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,4096,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,6144,0.18593688805898032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,5120,0.10177333487404717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,3584,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,4096,0.0817813343471951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,4096,0.12792355484432644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,3584,0.07295644283294678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,3072,0.012341332932313284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,3584,0.1141866710450914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,2560,0.01036533299419615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,3072,0.06360799736446805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,3072,0.10150400135252212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,2048,0.009343999955389235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,2560,0.052990221314960055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,2560,0.08796533611085679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,1536,0.008023999631404877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,2048,0.04366133279270596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,2048,0.07367377811008029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,1024,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,1536,0.033779554896884494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,1536,0.059361775716145836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,5120,0.15448444419436985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,768,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,1024,0.026138666603300307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,512,0.00489155575633049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,1024,0.0463084446059333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,768,0.021612443857722815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,768,0.03976088762283325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,256,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,512,0.018544000056054857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,512,0.03428533342149522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,128,0.0038399998512532976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,256,0.028075555960337322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,256,0.015134221977657743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,128,0.013691555294725629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,1024,128,0.023751111494170293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,64,0.004559111264016894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,1024,32,0.004868444469239977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,65536,0.16707821687062582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,32,0.01347644461525811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,1024,64,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,51200,0.12897777557373047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,65536,1.264068391587999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,51200,0.9899626837836372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,16384,0.04571200079388089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,16384,0.3260257773929172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,65536,1.7283795674641926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,12288,0.03598755598068237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,12288,0.2472675641377767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,16384,0.4503013292948405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,10240,0.028725332683987085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,51200,1.356610616048177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,12288,0.3430231147342258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,10240,0.20741332901848686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,8192,0.02436711059676276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,10240,0.28969155417548287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,7168,0.02109777761830224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,8192,0.16533954938252768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,8192,0.23388266563415527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,6144,0.018423999349276226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,7168,0.1434426705042521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,7168,0.20224355326758492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,5120,0.015962666935390897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,6144,0.12022222412957086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,6144,0.17239911026424834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,4096,0.014018666413095264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,5120,0.14401333861880833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,4096,0.07941155301200019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,4096,0.11860622300042047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,3584,0.012378666963842181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,3584,0.07045333253012763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,3584,0.10616355472140843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,3072,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,3072,0.06162133481767443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,3072,0.09344533416959976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,2560,0.009415999882751042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,2560,0.050516443120108716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,5120,0.0987466639942593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,2560,0.0819831093152364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,2048,0.008282666405042013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,1536,0.007280888656775157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,2048,0.04194844431347317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,2048,0.06913777854707506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,1536,0.03233688738611009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,1024,0.005913777897755305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,1536,0.054839111036724515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,1024,0.04371466570430332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,768,0.021347555849287245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,768,0.03756711218092177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,512,0.004579555657174853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,512,0.017901332841979135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,512,0.03289244572321574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,256,0.004311999926964442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,768,0.005594666633341048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,256,0.014483556151390076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,128,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,256,0.02642044425010681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,128,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,768,128,0.022651554809676275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,64,0.0035386664999855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,768,32,0.003607999947335985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,64,0.013076444466908773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,32,0.012723555167516073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,65536,0.12369778421190049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,768,1024,0.025428444147109985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,51200,0.09777333339055379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,65536,1.22416263156467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,51200,0.9592444101969401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,16384,0.033674667278925575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,65536,1.7279173533121746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,16384,0.31620176633199054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,12288,0.023944889505704243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,16384,0.4494320021735297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,12288,0.23964444796244302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,51200,1.3556079864501953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,12288,0.3429582118988037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,10240,0.20098843839433458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,8192,0.017265778448846605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,10240,0.28858666949801975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,8192,0.16114755471547446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,7168,0.014833778142929077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,8192,0.23400976922776964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,7168,0.13929333951738146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,10240,0.01994400057527754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,6144,0.013285333083735572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,7168,0.2015919950273302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,5120,0.011737777955002256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,6144,0.11673422654469807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,6144,0.17207733790079752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,4096,0.010054222411579555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,5120,0.09516800112194485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,5120,0.14474488629235163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,3584,0.009374221993817223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,4096,0.07661510838402642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,4096,0.11784088611602783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,3072,0.008347555167145198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,3584,0.06802044312159221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,3584,0.10628444618648952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,2560,0.007797333101431529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,3072,0.09329422314961751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,2048,0.006603555546866522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,2560,0.04921333326233757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,2048,0.0392524434460534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,2048,0.06709777646594577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,2560,0.08263555500242445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,1536,0.005980444451173146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,1536,0.05482044484880236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,1536,0.03150666753451029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,1024,0.004951999833186467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,3072,0.058697779973347984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,768,0.00462755560874939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,1024,0.04355022311210632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,1024,0.024880889389250014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,512,0.004268444246715969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,768,0.0370782216389974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,512,0.017613333132531907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,512,0.03247022297647264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,256,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,256,0.013849777479966482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,256,0.02725511127048069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,128,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,512,128,0.022700443863868713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,64,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,128,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,512,32,0.003334222154484855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,768,0.02069511181778378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,64,0.012477333347002665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,65536,0.07715911335415311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,512,32,0.012689777546458773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,51200,0.06078755855560303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,65536,1.1858844757080078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,51200,0.9298551347520617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,16384,0.024301333559883967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,65536,1.7163475884331598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,16384,0.3061270978715685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,12288,0.019283556275897555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,16384,0.44660091400146484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,12288,0.2327280044555664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,51200,1.3469875123765733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,10240,0.018702222241295707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,12288,0.34009777175055605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,8192,0.015800888339678448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,10240,0.2871137724982368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,8192,0.15577156013912624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,7168,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,8192,0.2319866551293267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,7168,0.13399822182125515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,6144,0.01221244451072481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,7168,0.20000710752275255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,6144,0.11322666539086236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,5120,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,10240,0.19450399610731336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,6144,0.1710604429244995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,4096,0.00793599999613232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,5120,0.09229422277874416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,5120,0.14309688409169516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,3584,0.007599111232492659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,4096,0.07500711414549086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,4096,0.11694310771094428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,3072,0.006991110742092133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,3584,0.06546133094363742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,3584,0.1047217779689365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,2560,0.006230222268237009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,3072,0.056652446587880455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,2048,0.00564177789621883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,2560,0.04660888844066196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,2560,0.08087199926376343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,1536,0.005117333183685939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,2048,0.03727022144529555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,2048,0.06578755378723145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,1024,0.004530666602982415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,1536,0.030382222599453394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,1536,0.0541013346778022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,768,0.004212444441186057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,1024,0.023989333046807185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,1024,0.04285066657596164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,512,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,768,0.036836445331573486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,768,0.019911999503771465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,256,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,512,0.016846223009957206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,512,0.03222399950027466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,128,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,256,0.013487999637921652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,256,0.025776000486479864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,64,0.0031679999083280563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,128,0.012235555383894177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,3072,0.09292711151970758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,256,32,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,64,0.01204533295498954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,65536,0.06106044186486138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,256,32,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,51200,0.050583998362223305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,256,128,0.02195466723706987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,65536,1.1487803988986545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,16384,0.015618667006492615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,51200,0.900511105855306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,16384,0.29712356461419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,12288,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,16384,0.44745424058702254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,12288,0.22533867094251844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,51200,1.3455244700113933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,10240,0.011896889242861005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,12288,0.34078666898939347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,10240,0.1891582277086046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,65536,1.7162791358100042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,8192,0.01075644459989336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,10240,0.28695556852552623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,7168,0.010400888820489248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,8192,0.15093866984049478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,8192,0.23132710986667207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,6144,0.010075555907355415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,7168,0.13026666641235352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,7168,0.1998951170179579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,6144,0.10919911331600612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,5120,0.00907199995385276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,5120,0.08905955817964341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,6144,0.17028000619676378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,5120,0.14247466458214655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,4096,0.007638221813572778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,3584,0.007304000357786815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,4096,0.11648000611199273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,4096,0.07228977812661065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,3584,0.06372444497214423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,3072,0.00675999994079272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,3584,0.10467377636167739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,2560,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,3072,0.05431288811895582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,3072,0.09313778082529704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,2048,0.005365333209435145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,2560,0.07981244723002116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,2048,0.06554400258594088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,1536,0.004869333157936732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,2048,0.037280887365341187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,1536,0.029779556724760268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,1024,0.004119111018048393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,1536,0.054529776175816856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,1024,0.023031999667485554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,768,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,1024,0.043219556411107383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,768,0.01957244508796268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,2560,0.04453777935769823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,512,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,768,0.03713422351413303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,256,0.003286222202910317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,512,0.016501333978441026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,512,0.032167110178205706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,128,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,256,0.025777777036031086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,256,0.013184000220563678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,128,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2560,128,128,0.02166666587193807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,32,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,64,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,65536,0.058557331562042236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,128,32,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,51200,0.047800001170900136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,128,64,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,16384,0.017880888448821176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,51200,0.899137814839681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,65536,1.1477777693006728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,12288,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,10240,0.01309511148267322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,16384,0.29716266526116264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,12288,0.22520266638861763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,8192,0.011873777541849347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,10240,0.18912266360388863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,7168,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,6144,0.01014755583471722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,8192,0.1508044401804606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,5120,0.008325333396593729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,7168,0.12971111138661703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,4096,0.007288888924651676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,6144,0.10956711239284939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,3584,0.006950221955776215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,5120,0.08894755442937215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,3072,0.006262222097979651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,4096,0.07192977931764391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,2560,0.005920889063013925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,3584,0.0638435549206204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,2048,0.005164444446563721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,1536,0.004679999831649992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,2560,0.044532444741990834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,1024,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,2048,0.037434667348861694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,1024,0.022983110613293115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,768,0.003952888978852166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,512,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,768,0.01928444372283088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,256,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,512,0.01663822266790602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,128,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,256,0.013065777719020844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,64,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,128,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,64,32,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,64,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,3072,0.05384977658589681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,65536,0.05743022097481621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,51200,0.046683556503719754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,32,0.011592888997660743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,64,1536,0.02941333254178365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,16384,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,16384,0.2969671090443929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,12288,0.00977600034740236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,51200,0.8991617626614041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,10240,0.009306666751702627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,65536,1.1481111314561632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,8192,0.008375110725561777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,12288,0.22495023409525552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,7168,0.00794400026400884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,10240,0.189082662264506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,6144,0.007694222033023834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,8192,0.15072977542877197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,5120,0.007319110963079665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,7168,0.1299591064453125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,4096,0.008292444050312042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,6144,0.10944533348083496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,3584,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,5120,0.08887822098202175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,3072,0.0069511110583941145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,4096,0.07087555858823988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,2560,0.006269333263238271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,3584,0.0634017785390218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,3072,0.05437866515583462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,2048,0.0052471111218134565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,1536,0.00453955555955569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,2560,0.04441510968738132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,1024,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,2048,0.03563466668128967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,768,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,1536,0.029502222935358684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,512,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,1024,0.02299555473857456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,256,0.002971555623743269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,768,0.01925244430700938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,512,0.016558221644825406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,256,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,64,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,64,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,32,0.0029315555261241067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,32,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2560,32,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2560,32,128,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,16384,2.4155040317111545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,16384,3.3841270870632596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,51200,7.330502827962239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,12288,1.786217795477973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,10240,1.4823840459187825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,12288,2.341783947414822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,10240,1.9167298210991754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,8192,1.19159730275472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,8192,1.5007751252916124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,51200,11.629714965820312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,7168,1.0764399634467232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,6144,0.9052026536729602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,7168,1.2902498245239258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,5120,0.7731715308295356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,6144,1.0972213745117188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,4096,0.6353511280483669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,5120,0.9087279637654623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,3584,0.547469351026747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,4096,0.7207715776231555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,3072,0.4790310859680176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,3584,0.634876463148329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,2560,0.38315288225809735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,3072,0.5454071362813314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,2048,0.30625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,2560,0.46384355756971574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,1536,0.2625262207455105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,1024,0.18938667244381377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,1536,0.29972177081637913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,768,0.15161777867211237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,2048,0.38086666001213926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,512,0.11203022797902425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,1024,0.22709245151943633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,256,0.07912621895472209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,768,0.18195288711123994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,128,0.05995821952819824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,64,0.05139822098943922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,128,0.1318302286995782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,256,0.1380293369293213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,65536,32,0.053780443138546415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,64,0.13345777988433838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,32,0.1341973278257582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,65536,512,0.15508000055948892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,51200,5.832864973280166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,65536,7.373219807942708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,16384,1.8553155263264973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,12288,1.5070230695936415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,51200,9.201417711046007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,16384,2.6813697814941406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,12288,1.8656959533691406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,10240,1.2084471384684246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,65536,11.806792365180122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,10240,1.5248070822821722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,8192,0.9301928414238824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,7168,0.8160782390170627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,8192,1.1855590608384874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,7168,1.0313528908623588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,6144,0.7461831304762098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,5120,0.6085937817891439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,6144,0.8777830865648059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,5120,0.7305804358588325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,4096,0.49887731340196395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,3584,0.4249635537465413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,4096,0.576416015625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,3584,0.5298737949795193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,3072,0.37187732590569395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,2560,0.3108595477210151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,3072,0.43806134329901797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,2048,0.27299287584092885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,2560,0.37173067198859316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,2048,0.3053022225697835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,1536,0.21048711405860054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,1024,0.14832354916466609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,1536,0.24168801307678223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,768,0.12030577659606934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,512,0.08963200118806626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,1024,0.18345599704318574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,768,0.14663555887010363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,256,0.0624177787039015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,128,0.05073777834574381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,512,0.12475200494130452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,64,0.04456444581349691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,256,0.11123289002312554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,51200,32,0.047190222475263804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,64,0.10734399822023179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,128,0.10617333650588989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,51200,32,0.10838755634095933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,51200,1.7916622161865234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,65536,2.2778826819525824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,16384,0.5869759983486599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,16384,0.9846604665120443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,12288,0.44132354524400497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,65536,4.383565266927083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,10240,0.4063742160797119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,12288,0.7138560083177355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,8192,0.33385777473449707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,10240,0.5861600240071615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,7168,0.29693510797288686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,51200,3.428113725450304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,8192,0.4647093349032932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,6144,0.2649928993648953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,5120,0.20381422837575278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,7168,0.4038320117526584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,4096,0.157916439904107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,6144,0.3440115451812744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,5120,0.2918764485253228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,3584,0.13932355244954428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,3072,0.12204355663723415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,4096,0.22820001178317598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,3584,0.20141243934631348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,2560,0.10941244496239556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,2048,0.08491466442743938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,3072,0.17548710770077178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,1536,0.06552177667617798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,1024,0.04886577857865227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,2560,0.1507022248374091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,1536,0.09910577535629272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,1024,0.07515199979146321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,512,0.02938844429122077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,768,0.06194133228725857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,256,0.021958221991856892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,512,0.05173688795831469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,128,0.018244443668259513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,256,0.04558222161398994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,64,0.015971556305885315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,128,0.042816887299219765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,32,0.016132444143295288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,64,0.04146044452985128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,2048,0.12427822748819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,16384,32,0.0418524444103241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,16384,768,0.040208889378441706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,51200,1.333563592698839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,65536,1.673633787367079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,16384,0.4306320084465875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,12288,0.3311599890391032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,16384,0.7962337599860297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,10240,0.3057120111253527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,12288,0.5751146740383571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,51200,2.7552729712592234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,8192,0.24686222606235078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,10240,0.5023137728373209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,7168,0.20882756180233428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,65536,3.5582792494032116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,8192,0.3786986668904622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,7168,0.32951998710632324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,5120,0.1564071045981513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,6144,0.2796613375345866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,4096,0.12303733825683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,5120,0.2330231136745877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,3584,0.11212889353434245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,4096,0.18685955471462676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,3072,0.09405244721306695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,2560,0.08155377705891927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,3072,0.14430667294396296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,2048,0.06961155600017972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,6144,0.18227999740176729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,1536,0.051560001240836255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,2048,0.10269688897662693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,2560,0.12464800145890977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,1536,0.0823520024617513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,1024,0.03768622212939792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,768,0.030204445123672485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,1024,0.06202489137649536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,512,0.02215733296341366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,3584,0.16556889481014675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,256,0.017468444175190397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,768,0.051312890317704946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,512,0.041477332512537636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,128,0.014550222290886773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,64,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,256,0.035007112556033664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,128,0.0332640012105306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,12288,32,0.012489777472284106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,32,0.032300445768568255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,12288,64,0.033082667324278094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,65536,1.5356666776869032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,51200,1.2483030954996746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,16384,0.41052622265285915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,16384,0.7145519786410861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,65536,3.1211217244466147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,12288,0.5196835729810927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,51200,2.484776814778646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,10240,0.26417067315843373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,12288,0.3048622343275282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,8192,0.2090515560574002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,10240,0.4255519972907172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,7168,0.1816106637318929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,6144,0.1513706710603502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,7168,0.29271554946899414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,8192,0.3362906773885091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,5120,0.1351111067665948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,6144,0.24862578180101183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,4096,0.10771110985014175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,3584,0.08896178007125854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,4096,0.16704977883232963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,5120,0.2072168853547838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,3072,0.07922666602664523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,3584,0.147434671719869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,2560,0.06502755482991536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,2048,0.05522133244408501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,1536,0.04503200120396084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,2048,0.09227822224299113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,2560,0.11093422439363267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,3072,0.12924799654218885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,1536,0.07333511114120483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,768,0.025755556093321905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,512,0.019166222876972623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,768,0.04525688952869839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,256,0.014596444037225513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,256,0.03033777740266588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,512,0.03680355681313409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,128,0.012806221842765808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,64,0.011448889142937131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,64,0.028910222980711196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,32,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,128,0.029346668057971533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,10240,1024,0.03129244512981839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,32,0.028509332074059382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,10240,1024,0.05493155452940199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,51200,0.9162311553955078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,65536,1.2190266715155709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,16384,0.310007996029324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,12288,0.22751111454433867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,16384,0.6073137919108073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,10240,0.20424355400933158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,12288,0.46079646216498477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,51200,2.0922658708360458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,8192,0.17904888259039986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,7168,0.14524622758229574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,65536,2.6476044125027127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,8192,0.2936462296379937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,6144,0.12156177891625299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,10240,0.37197422981262207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,5120,0.10393689076105754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,7168,0.25957955254448783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,4096,0.08162933588027954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,6144,0.21684355205959746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,3584,0.07266400257746379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,3072,0.062302225165896945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,4096,0.1451573371887207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,5120,0.18096800645192465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,2560,0.05379644367429945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,3584,0.12936177518632677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,2048,0.04467555549409655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,3072,0.11343377166324192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,1536,0.037963555918799505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,1024,0.026325333449575637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,2048,0.08110666937298246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,768,0.02160355614291297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,2560,0.09836266438166301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,1536,0.06430221928490533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,1024,0.04795999990569221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,512,0.015998222761683993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,256,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,768,0.03982844286494785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,128,0.010400888820489248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,512,0.03161155515246921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,64,0.00906666699382994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,128,0.025388444463411968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,64,0.025798221429189045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,256,0.027216888136333887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,8192,32,0.009397333694828881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,8192,32,0.025415110919210646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,51200,0.8152346611022949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,65536,1.0702898237440321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,65536,2.4705191718207464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,16384,0.2701031102074517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,12288,0.19915644327799478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,51200,1.909163581000434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,16384,0.5609368748135036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,10240,0.17028355598449707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,8192,0.13698755370246038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,12288,0.41428621610005695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,10240,0.3454079892900255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,7168,0.12470132774776882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,8192,0.2727697690327962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,6144,0.10957333114412095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,7168,0.23738132582770455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,5120,0.0908471080991957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,4096,0.07452799876530965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,6144,0.20156533188290068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,3584,0.06522755490409003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,4096,0.1347377830081516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,3584,0.11960089206695557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,3072,0.05741777684953478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,2560,0.049080888430277504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,3072,0.10603288809458415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,2048,0.03914133376545376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,2560,0.09170489178763495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,2048,0.07582488987180921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,1536,0.03090755475891961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,1024,0.02253866692384084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,5120,0.16710755560133192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,768,0.01843377782238854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,1536,0.05977955791685316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,512,0.013261333107948303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,1024,0.04469244347678291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,768,0.03699288765589396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,512,0.029337777031792536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,128,0.009365333451165093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,256,0.02476177778508928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,64,0.008282666405042013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,128,0.023805333508385554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,7168,32,0.008466666771305932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,64,0.023697778582572937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,7168,32,0.023407111565272015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,51200,0.7464684380425347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,65536,0.9708400302463107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,16384,0.25914133919609916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,12288,0.19922310776180693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,16384,0.5185128847757975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,10240,0.15498488479190403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,12288,0.38243643442789715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,65536,2.2433670891655817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,8192,0.13563556141323513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,10240,0.3193742169274224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,7168,0.10905066463682388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,6144,0.09870400031407674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,51200,1.7387075424194336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,7168,0.2196364402770996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,5120,0.07819110817379422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,4096,0.06212088796827528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,5120,0.15444355540805393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,6144,0.18645777967241076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,3584,0.057123555077446826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,3072,0.05068177647060818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,3584,0.111026664574941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,3072,0.0971191128094991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,2560,0.04192177785767449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,2048,0.034541332059436373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,2560,0.0844835572772556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,1536,0.026199110680156287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,2048,0.07042044401168823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,1024,0.019504000743230183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,1536,0.055135111014048256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,768,0.01607466737429301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,4096,0.1250577767690023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,512,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,768,0.034635557068718806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,1024,0.04119999872313605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,256,0.011649777491887411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,128,0.008345777789751688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,512,0.027114666170544092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,64,0.007304888632562425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,256,0.023384888966878254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,128,0.02200355629126231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,6144,32,0.007691555553012424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,64,0.021980444590250652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,8192,0.2528293397691515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,6144,32,0.021710221966107685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,51200,0.6217279964023167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,65536,0.807005352444119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,16384,0.21630311012268066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,12288,0.1627688937717014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,16384,0.5020177629258897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,12288,0.3589351177215576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,10240,0.13524444897969565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,51200,1.5814773771497939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,8192,0.11992355187733968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,65536,2.0032374064127603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,10240,0.293525324927436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,7168,0.09778755240970188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,8192,0.23264622688293457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,7168,0.20150132973988852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,5120,0.06950133376651339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,6144,0.1708666616015964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,4096,0.05934222539265951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,5120,0.14170400301615396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,3584,0.05045510994063484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,4096,0.11468799908955891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,3072,0.04356355468432108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,3584,0.10211288928985596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,2560,0.03685866792996725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,3072,0.08915111091401841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,2048,0.028014222780863445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,6144,0.08286933104197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,2560,0.0746506651242574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,1536,0.02240266733699375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,1024,0.016456888781653512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,2048,0.06449155675040351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,1536,0.050423999627431236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,768,0.01425333321094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,512,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,1024,0.03763466742303636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,256,0.00833511104186376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,768,0.031679110394583807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,128,0.007314666277832455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,512,0.025046222739749487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,64,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,256,0.02139644490347968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,5120,32,0.006659555352396435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,128,0.020255110330051847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,64,0.020253333780500624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,5120,32,0.020253333780500624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,65536,0.805348449283176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,51200,0.629067579905192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,16384,0.20871022012498644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,16384,0.42889067861768937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,12288,0.15782044993506536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,65536,1.804473876953125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,10240,0.11502488454182942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,12288,0.32092176543341744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,51200,1.4147013558281794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,8192,0.08976622422536214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,10240,0.2679742177327474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,7168,0.0797777771949768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,6144,0.06911022133297391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,8192,0.21209422747294107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,5120,0.05832711193296644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,7168,0.18375021881527373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,4096,0.048173334863450795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,6144,0.15566844410366482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,3584,0.04186933239301046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,5120,0.12973244984944662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,3072,0.032799111472235784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,4096,0.10437955458958943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,2560,0.028263999356163874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,3584,0.09318133195241292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,2048,0.023580445183648005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,3072,0.08097244633568658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,1536,0.0196782218085395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,2560,0.06795199712117513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,1024,0.01440533333354526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,2048,0.055868446826934814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,768,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,1536,0.04557955596182081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,512,0.009027555584907532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,1024,0.0329386658138699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,256,0.006980444822046492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,768,0.028786665863460962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,128,0.006245333287451003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,512,0.02269688910908169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,64,0.005946666830115848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,256,0.019687111179033916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,4096,32,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,64,0.01854044364558326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,32,0.018259555101394653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,51200,0.4155742327372233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,65536,0.5235662460327148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,16384,0.1390168931749132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,4096,128,0.018814222680197824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,12288,0.10318399800194634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,16384,0.4070320129394531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,10240,0.08640711175070868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,12288,0.3058862156338162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,65536,1.6913688447740343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,8192,0.08415644698672825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,10240,0.25503378444247776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,51200,1.3252106772528756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,7168,0.07268355290095012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,6144,0.06260533465279473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,8192,0.20175111293792725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,5120,0.05241511265436808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,4096,0.04179999894566006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,7168,0.17486045095655653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,6144,0.14786222245958117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,5120,0.12350755267673069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,3584,0.036369777388042875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,4096,0.09954666429095799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,3072,0.031094223260879517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,3584,0.08850489060084026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,2560,0.02679288884003957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,2048,0.02201244400607215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,3072,0.07685333490371704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,2560,0.06502399841944377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,1536,0.017992888887723286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,1024,0.013087111214796701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,2048,0.05409155620468987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,1536,0.04282933473587036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,768,0.011178666518794166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,512,0.00847999999920527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,1024,0.03197333216667175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,256,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,768,0.027458666099442378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,128,0.0063075555695427795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,512,0.02239199976126353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,256,0.018917333748605516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,64,0.005700444595681296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3584,32,0.005934222290913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,128,0.01793688866827223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,64,0.01761244402991401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3584,32,0.018283555905024212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,65536,0.4682337972852919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,51200,0.3657982349395752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,16384,0.12388355202145046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,16384,0.3845795525444879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,12288,0.09364622169070774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,65536,1.5850168863932292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,51200,1.23634582095676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,10240,0.07947822411855061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,8192,0.0654675563176473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,12288,0.2906320095062256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,7168,0.058824002742767334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,10240,0.2412346733940972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,6144,0.04936977889802721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,8192,0.19185245037078857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,5120,0.04171289006868998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,7168,0.16631200578477648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,4096,0.0339937773015764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,6144,0.14038222365909153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,5120,0.1168702178531223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,3584,0.030256887276967365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,3072,0.026730666557947796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,3584,0.08394933409161037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,4096,0.09519822067684597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,3072,0.07287377781338163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,2560,0.023581332630581323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,2048,0.01984355515903897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,1536,0.015955555770132277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,2560,0.0634124411476983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,2048,0.05132800009515551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,1536,0.040454222096337214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,1024,0.01235466698805491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,768,0.010031110710567897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,512,0.007675555845101674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,1024,0.030224889516830444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,256,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,768,0.02625333269437154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,512,0.021003555920388963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,128,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,256,0.017982222967677645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,128,0.016920000314712524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,32,0.005575110928879843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,64,0.016594666573736403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,3072,32,0.016878222425778706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,65536,0.4322071075439453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,51200,0.337860451804267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,3072,64,0.0053022222386466126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,16384,0.09370844231711493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,65536,1.471760008070204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,12288,0.07480000125037299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,10240,0.07097333007388644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,16384,0.3648391034868028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,51200,1.158323605855306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,8192,0.055339554945627846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,7168,0.044963555203543774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,12288,0.2744693226284451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,6144,0.04136000076929728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,8192,0.18196800020005968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,5120,0.03515200151337518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,7168,0.15720621744791666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,6144,0.13264977931976318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,4096,0.02828888760672675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,3584,0.02552177839808994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,5120,0.11005600293477376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,3072,0.023016000787417095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,4096,0.0893439981672499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,2560,0.02045866681469811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,3584,0.08054400152630277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,2048,0.016100444727473788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,3072,0.07020622491836548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,1536,0.013339555925793119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,2560,0.05898577637142605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,1024,0.009666666388511658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,2048,0.04885244369506836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,1536,0.03750133183267381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,768,0.008389333056079017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,1024,0.02844088938501146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,512,0.006934222247865465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,768,0.024714666936132643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,256,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,512,0.019959110352728102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,128,0.005103999955786599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,256,0.01718755563100179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,64,0.004616888860861461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,128,0.016161777906947665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2560,32,0.00489066665371259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,64,0.015847999188635085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,32,0.015820443630218506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,65536,0.3416755464341905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2560,10240,0.22857599788241914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,51200,0.26943733957078725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,16384,0.09117511245939468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,16384,0.3424239953358968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,12288,0.07114222314622667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,65536,1.3469911151462133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,51200,1.0585270987616646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,10240,0.06062044699986776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,8192,0.0476399991247389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,12288,0.25868622461954754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,7168,0.041040887435277305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,10240,0.215575999683804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,8192,0.17106578085157606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,6144,0.033949332104788885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,5120,0.02851111027929518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,7168,0.14744977156321207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,4096,0.024346666203604803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,6144,0.12495556142595078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,3584,0.021811554829279583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,5120,0.10434044731987847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,3072,0.018375110295083787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,4096,0.08436799711651272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,2560,0.015994666351212394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,3584,0.07505333423614502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,3072,0.06529244449403551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,2560,0.05451377895143297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,1536,0.011407111254003314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,2048,0.04514044523239136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,1024,0.00871377769443724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,1536,0.0356391105386946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,768,0.007674666742483775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,1024,0.027111111415757075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,512,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,768,0.02307111190425025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,256,0.005679111099905438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,512,0.018985778093338013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,128,0.004929777648713854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,256,0.01588622232278188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,64,0.004602666530344221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,128,0.015512888630231222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,2048,0.014170666535695394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,2048,32,0.004587555511130227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,64,0.014993778533405729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,2048,32,0.014862222803963555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,65536,0.2622702121734619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,51200,0.2060177723566691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,16384,0.0722844468222724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,16384,0.32145688268873424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,12288,0.0541013346778022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,51200,0.9838533401489258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,65536,1.251736005147298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,10240,0.04565511147181193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,8192,0.03730400072203742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,12288,0.24308978186713326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,7168,0.031403554810418025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,10240,0.20348710483974883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,8192,0.16018488672044542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,6144,0.027832001447677612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,7168,0.1386257807413737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,5120,0.024175110790464614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,4096,0.020424000091022916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,6144,0.11727200614081489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,5120,0.09728799925910102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,3584,0.01829777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,3072,0.016166221764352586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,4096,0.07878133323457506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,2560,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,3584,0.07039378086725871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,2048,0.01255644443962309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,2560,0.05092799994680616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,1536,0.009451555709044138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,2048,0.04202222161822849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,1024,0.007298666569921706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,1536,0.033667554457982383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,768,0.006637333167923822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,1024,0.025076443950335186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,512,0.00563822231358952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,768,0.022063111265500385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,256,0.00462666650613149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,512,0.01794666714138455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,128,0.004386666748258802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,256,0.015211555692884656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,64,0.004280888785918553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,128,0.014449778530332776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1536,32,0.0041982221106688184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,3072,0.06096533272001478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,64,0.014326221413082547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1536,32,0.014113777213626437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,65536,0.19717866844601104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,51200,0.15959733062320286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,51200,0.9139297273423937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,16384,0.04810133245256212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,16384,0.30123644404941136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,12288,0.0347031123108334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,12288,0.2277777724795871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,10240,0.0287715560860104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,65536,1.1625066333346896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,8192,0.02420266634888119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,10240,0.19035644001430937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,7168,0.019761777586407132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,8192,0.15078311496310765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,6144,0.018784888916545443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,7168,0.12978577613830566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,5120,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,6144,0.10897955629560684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,4096,0.013501333693663279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,5120,0.09081866343816121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,3584,0.011889778077602386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,4096,0.07367644707361858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,3072,0.010804444551467896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,3584,0.06525778108172946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,2560,0.009855111440022787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,3072,0.056398219532436795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,2048,0.008343111309740279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,2560,0.04713066750102573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,1536,0.007248889240953658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,1536,0.030893332428402368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,1024,0.005913777897755305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,1024,0.02401155564520094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,768,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,768,0.020613332589467365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,512,0.004968889057636261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,2048,0.03778488768471612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,512,0.016912000046836007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,256,0.0041306668685542215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,128,0.003912888881233004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,256,0.014520888527234396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,64,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,128,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,1024,32,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,64,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,1024,32,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,65536,0.15489510695139566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,51200,0.12514755460951063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,16384,0.040693332751592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,16384,0.2910319964090983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,12288,0.02950577934583028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,51200,0.8830933570861816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,65536,1.124201774597168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,12288,0.22031822469499376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,10240,0.0236124445994695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,8192,0.019190222024917603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,7168,0.01610666679011451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,10240,0.183643552992079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,6144,0.014383999837769402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,8192,0.14563466442955866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,7168,0.12544533941480848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,5120,0.012477333347002665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,4096,0.010589333044158088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,3584,0.010090666512648264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,6144,0.10606577661302354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,3072,0.008740444150235917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,5120,0.08753955364227295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,4096,0.07170755333370633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,3584,0.0630177789264255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,2048,0.007150222029950883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,3072,0.0540888872411516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,1536,0.005952000204059813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,2560,0.045217778947618276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,1024,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,2048,0.03698222173584832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,768,0.004744889007674323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,1536,0.029912001556820337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,512,0.004541333350870344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,1024,0.023402666052182514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,768,0.019911110401153564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,256,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,512,0.01648533344268799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,128,0.00360977773865064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,64,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,256,0.013531555732091268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,32,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,128,0.01308444473478529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,64,0.013314666847387949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,65536,0.12420711252424453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,768,2560,0.008400888906584846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,51200,0.09366933504740398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,16384,0.03182755576239692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,16384,0.280694219801161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,12288,0.022452443838119507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,51200,0.8468764623006185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,12288,0.21286488903893364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,10240,0.01904000010755327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,65536,1.0834177864922416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,8192,0.016140444411171805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,7168,0.01421866648726993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,768,32,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,10240,0.17727111445532906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,6144,0.013024888932704926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,8192,0.14044888814290366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,7168,0.12104533778296576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,5120,0.011133333047231039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,4096,0.009733333355850643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,6144,0.10180622339248657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,5120,0.0843013326327006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,3584,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,3072,0.008344000412358178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,4096,0.0690408878856235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,3584,0.060643553733825684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,3072,0.05151377783881294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,2048,0.006595555692911148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,2560,0.042916443612840444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,1536,0.005667555663320754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,2048,0.0351528889603085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,1024,0.004592000196377437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,1536,0.02853155467245314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,768,0.0046364445653226645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,1024,0.02237422267595927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,512,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,768,0.019360888335439894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,256,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,512,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,128,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,2560,0.007340444458855524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,256,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,64,0.0033155555526415506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,128,0.0127688886390792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,512,32,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,64,0.012144889268610211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,65536,0.0702257752418518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,512,32,0.012152888708644442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,51200,0.05679821968078613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,16384,0.020918221937285528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,16384,0.2705155478583442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,12288,0.02127644419670105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,51200,0.8147866461012098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,65536,1.0398648579915364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,10240,0.014706666270891825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,8192,0.01240533341964086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,12288,0.20533511373731825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,10240,0.17190399434831408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,7168,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,8192,0.13494133949279785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,6144,0.010791111323568555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,5120,0.0090524446633127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,7168,0.11646578047010635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,6144,0.09779110882017349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,4096,0.008096888661384583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,3584,0.007312888900438945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,5120,0.08150044414732191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,3072,0.006709333509206772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,4096,0.06544444296095106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,2560,0.005966222120655908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,3584,0.05788266658782959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,2048,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,3072,0.049601776732338794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,2560,0.04157866703139411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,1536,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,1024,0.004260444392760594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,2048,0.03366133239534166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,768,0.003940444439649582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,1536,0.027752000424597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,512,0.00361333332127995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,1024,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,768,0.018969777557584975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,256,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,128,0.0032026666320032547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,256,0.01275288893116845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,64,0.002986666642957263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,128,0.012472000386979846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,256,32,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,64,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,65536,0.059024002816942006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,32,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,51200,0.047689778937233814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,256,512,0.01552177800072564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,16384,0.016128889388508268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,65536,1.039615101284451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,51200,0.8146746423509387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,12288,0.013526221944226159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,16384,0.2697715494367811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,10240,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,8192,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,7168,0.010724444356229572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,10240,0.17093066374460855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,12288,0.20529066191779244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,6144,0.009684444301658207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,8192,0.13549333148532444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,5120,0.008667555948098501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,4096,0.007288888924651676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,7168,0.11619822184244792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,6144,0.097543113761478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,3584,0.00699733363257514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,3072,0.006341333190600078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,4096,0.06501333581076728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,2560,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,3584,0.05756622552871704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,3072,0.04853244291411506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,2560,0.04036622246106466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,2048,0.005223111146026188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,1536,0.004906666775544484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,2048,0.03329688972897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,1024,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,1536,0.027479110492600337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,768,0.003932444585694207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,1024,0.02165066699186961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,512,0.00351555562681622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,768,0.018568888306617737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,256,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,5120,0.08015555805630155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,512,0.01516533394654592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,128,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,64,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,256,0.01276000009642707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,128,32,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,128,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,64,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,65536,0.051402668158213295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,128,32,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,51200,0.04216444492340088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,16384,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,16384,0.269540442360772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,12288,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,51200,0.8136373096042209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,10240,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,12288,0.20460977819230822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,65536,1.0389297273423936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,8192,0.010718222293588849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,7168,0.009891555541091496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,10240,0.17113333278232148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,6144,0.009663110805882348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,8192,0.1352648867501153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,7168,0.11597955226898193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,5120,0.008997333546479544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,4096,0.007634667058785756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,6144,0.09737600220574273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,3584,0.007279111279381647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,5120,0.08023555411232842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,4096,0.06477422184414335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,3072,0.006281777802440856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,2560,0.005904888941182031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,3584,0.05746666590372721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,2048,0.005219555563396878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,2560,0.040031111902660794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,1536,0.004538666870858935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,2048,0.03321866525544061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,1024,0.0041955556306574075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,1536,0.027450667487250432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,768,0.0038399998512532976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,1024,0.02126933303144243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,512,0.0035546666218174826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,768,0.018588445252842374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,256,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,512,0.015230221880806817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,256,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,64,0.002540444541308615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,128,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,64,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,64,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,3072,0.04831377665201823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,65536,0.052080889542897545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,64,32,0.011034666664070554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,51200,0.040483554204305015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,16384,0.011499555574523078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,16384,0.2692462338341607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,12288,0.009302222066455418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,51200,0.8136871125963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,10240,0.008624888956546783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,12288,0.2046417792638143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,65536,1.0385831197102864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,8192,0.007639110916190677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,7168,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,10240,0.1712017721600003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,8192,0.13524977366129556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,6144,0.0076142218377855085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,5120,0.017163554827372234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,6144,0.0967715581258138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,4096,0.01479111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,5120,0.08022844129138522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,3584,0.013092444174819522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,4096,0.06495110856162177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,3072,0.01164533363448249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,3072,0.04843822121620178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,2560,0.010306666294733683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,2560,0.04081511166360643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,2048,0.0052800000541739995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,1536,0.004582222137186262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,1536,0.027518222729365032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,1024,0.0041413332025210066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,3584,0.057421333260006376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,1024,0.0210479994614919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,2048,0.033914665381113686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,768,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,512,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,256,0.0032337777730491427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,768,0.018403554956118267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,512,0.01518311103185018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,128,0.0029120000286234748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,256,0.012795555922720166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,64,0.0029288888391521242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,128,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,7168,0.11555822690327962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2305,32,32,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,64,0.011805333197116852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2305,32,32,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,51200,6.931189219156901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,16384,2.450126224093967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,16384,3.251760058932834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,51200,11.198306613498263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,12288,1.6628541946411133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,51200,7.547518412272136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,16384,2.397407955593533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,12288,2.294893264770508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,10240,1.4989795684814453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,12288,1.8052950965033636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,8192,1.1975306404961479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,10240,1.8636514875623915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,10240,1.5248009363810222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,7168,1.1172951592339408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,8192,1.4646772808498805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,8192,1.2213253445095487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,7168,1.2556649314032662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,6144,0.8614880243937174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,7168,1.062837282816569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,5120,0.7168035507202148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,6144,1.065311113993327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,6144,0.9173875384860568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,4096,0.5743146472507054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,5120,0.8829831017388238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,5120,0.7761715783013238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,3584,0.5169564353095161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,4096,0.7039004431830512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,4096,0.6338506804572211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,3072,0.4578133159213596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,3584,0.6156009038289388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,3584,0.5633120006985134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,2560,0.3737315601772732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,3072,0.5335626602172852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,3072,0.4939902093675401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,2560,0.4548800256517198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,2048,0.2998746765984429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,2560,0.422560003068712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,1536,0.23894045088026258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,2048,0.37385421329074436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,2048,0.35495911704169375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,1024,0.16590221722920737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,1536,0.2952977816263835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,1536,0.28330932723151314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,768,0.13404533598158094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,1024,0.22357511520385742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,1024,0.2177893320719401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,512,0.09930666950013901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,768,0.17922488848368326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,256,0.06968444585800171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,768,0.1809555557039049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,512,0.1466639969083998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,512,0.15226756201850042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,128,0.059280885590447314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,256,0.11726844310760498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,256,0.13551466994815403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,64,0.05300355619854397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,65536,32,0.05651377969317966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,65536,128,0.0925439993540446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,128,0.12935377491845026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,64,0.13040444586012098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,65536,32,0.13158044550153944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,51200,5.312669542100695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,65536,6.868583255343967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,65536,11.311036851671007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,65536,7.860266791449653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,16384,1.7865280575222438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,51200,8.942242092556423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,16384,2.611312018500434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,12288,1.2674471537272136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,16384,1.960641860961914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,51200,6.183834499782986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,10240,1.1578470865885417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,12288,1.8236497243245442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,12288,1.4744257397121852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,10240,1.4876622094048395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,8192,1.1751848856608074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,10240,1.2358826531304252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,8192,0.9825688468085395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,8192,0.9958115683661567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,7168,0.782631132337782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,6144,0.679749329884847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,7168,1.004151980082194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,7168,0.8693040211995443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,6144,0.8592098024156359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,5120,0.5565937889946831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,5120,0.7086666425069174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,6144,0.7521920204162598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,4096,0.47658491134643555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,5120,0.6360684500800239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,4096,0.563002692328559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,3584,0.4244053363800049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,3584,0.4940257602267795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,4096,0.5220959981282552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,3072,0.34969332483079696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,3584,0.46099646886189777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,3072,0.42798222435845273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,2560,0.29373155699835884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,3072,0.41765867339240176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,2560,0.36516533957587344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,2048,0.24761509895324707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,2560,0.34560267130533856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,1536,0.18504267268710664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,2048,0.29974667231241864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,2048,0.29831910133361816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,1536,0.2397164503733317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,1024,0.14011555247836643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,1536,0.2335431045956082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,768,0.10018933481640285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,1024,0.1796791156133016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,1024,0.17986933390299478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,512,0.07283022006352742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,768,0.14468267228868273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,768,0.14980710877312556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,256,0.054478221469455294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,512,0.12219288614061143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,512,0.1225306722852919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,128,0.04655377732382881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,256,0.09735733270645142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,256,0.10924888981713189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,64,0.03949866692225138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,128,0.10397866699430679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,51200,32,0.040560000472598605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,51200,128,0.07743289073308308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,64,0.10488266415066189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,51200,32,0.10572444730334812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,65536,2.3139093187120223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,51200,1.8135448031955297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,65536,4.299773322211371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,16384,0.567662239074707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,65536,3.517688963148329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,16384,0.9628666771782769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,12288,0.42165154880947536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,51200,3.33935907151964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,16384,0.8965288798014323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,10240,0.3783911069234212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,12288,0.69559113184611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,12288,0.6798719829983182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,10240,0.5711475478278266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,8192,0.3094186782836914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,10240,0.5716924667358398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,51200,2.7665333218044705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,8192,0.4515546692742242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,7168,0.27256443765428334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,7168,0.3949893315633138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,6144,0.2413324515024821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,8192,0.45958222283257383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,7168,0.40098044607374406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,6144,0.33483288023206925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,5120,0.1880577802658081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,6144,0.34726399845547146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,4096,0.15347111225128174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,5120,0.27885601255628795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,5120,0.2922862105899387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,3584,0.13640711042616102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,4096,0.22285334269205728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,3072,0.12441333134969075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,3584,0.19731822278764513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,3584,0.2137653297848172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,4096,0.2394639915890164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,2560,0.09801955355538262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,3072,0.1722897821002536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,3072,0.18878755304548475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,2048,0.07998844650056627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,2560,0.16233332951863608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,2560,0.14878311422136095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,1536,0.06278577778074476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,2048,0.12264445092942979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,2048,0.13831378353966609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,1024,0.0457911127143436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,1536,0.11233154932657878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,1536,0.09826844268374973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,1024,0.07439644469155206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,768,0.038450668255488075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,512,0.028110222684012517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,768,0.07461155785454644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,768,0.06165066692564222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,512,0.06172355678346422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,256,0.020439111524158053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,256,0.04456622070736355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,256,0.050491554869545825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,128,0.018586667047606576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,128,0.04147288865513272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,128,0.04193511274125841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,64,0.01593155496650272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,16384,32,0.015524443652894763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,16384,1024,0.08730666504965888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,64,0.04160888989766439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,32,0.04046577877468533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,16384,512,0.05116622315512764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,51200,1.2986204359266493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,65536,1.798537784152561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,65536,3.406988355848524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,51200,2.689425786336263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,65536,3.0022159152560763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,16384,0.4180933369530572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,16384,0.7760684225294324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,12288,0.34238399399651426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,16384,0.7667022281222873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,12288,0.5975075297885472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,10240,0.2930080095926921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,12288,0.5811582141452366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,51200,2.34990480211046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,10240,0.48707734213935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,8192,0.2363057666354709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,10240,0.4892702102661133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,8192,0.3702133231692844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,7168,0.21441955036587187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,8192,0.39333687888251406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,7168,0.3228951030307346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,6144,0.17820355627271864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,7168,0.3871440092722575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,5120,0.1459991137186686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,6144,0.2742951181199816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,6144,0.2966017723083496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,5120,0.2285848988427056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,5120,0.250328885184394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,4096,0.12464443842569987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,3584,0.10548266437318589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,4096,0.18381244606441924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,4096,0.20622489187452528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,3072,0.09214488665262859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,3584,0.18361955218844941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,2560,0.07756977611117892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,3584,0.16321244504716662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,3072,0.14212444093492296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,3072,0.16179022524091932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,2048,0.06106488572226631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,2560,0.12238311767578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,1536,0.04858488837877909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,2048,0.10165155596203274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,2048,0.11914043956332737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,2560,0.140355560514662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,1024,0.0355368885729048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,1536,0.0812986691792806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,1536,0.09728178050782944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,768,0.027458666099442378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,1024,0.07576000028186373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,512,0.020597333709398907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,768,0.0494044456217024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,768,0.06549155712127686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,1024,0.06174311372968885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,256,0.016368000043763053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,512,0.03982399900754293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,128,0.013792888985739814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,512,0.054069333606296115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,256,0.043953776359558105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,64,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,128,0.032264888286590576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,12288,128,0.03738844394683838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,256,0.03563377923435635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,12288,32,0.0120319997270902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,64,0.0326257778538598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,12288,32,0.03221866819593642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,51200,1.228006256951226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,65536,1.5801297293768988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,65536,3.0523111555311413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,16384,0.40962934494018555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,51200,2.366141425238715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,65536,2.7422462039523654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,16384,0.6844017770555285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,12288,0.3225182162390815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,51200,2.1476283603244357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,16384,0.702546649509006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,12288,0.5016364521450466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,10240,0.24294845263163248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,8192,0.1893484459982978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,10240,0.41564533445570206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,10240,0.44867197672526044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,7168,0.16912266943189833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,8192,0.3302675618065728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,8192,0.3597617679172092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,6144,0.14945956071217856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,7168,0.2873377799987793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,12288,0.5828426678975424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,7168,0.31539644135369194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,6144,0.24392623371548125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,5120,0.1283253298865424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,6144,0.27035734388563365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,4096,0.10345511304007636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,5120,0.22796800401475695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,4096,0.16379466321733263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,3584,0.08833066622416179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,5120,0.20312889417012533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,4096,0.18892978297339547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,3072,0.07721955246395536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,3584,0.1449635558658176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,3584,0.16796710756089953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,3072,0.12712889247470432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,2560,0.06284444199668036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,2048,0.051402668158213295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,2560,0.10940977599885728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,3072,0.14954045083787706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,2560,0.1283644437789917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,1536,0.04252533449067009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,2048,0.09131733576456706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,1536,0.08966310818990071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,1536,0.07257066832648383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,2048,0.1105075544781155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,1024,0.029697779152128432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,768,0.023266666465335425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,1024,0.05451200074619717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,1024,0.07046577665540907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,768,0.044212443961037524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,512,0.018231110440360177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,768,0.06052622530195448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,256,0.014371555712487964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,512,0.03535111082924737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,512,0.05110044280687968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,256,0.03997600078582764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,128,0.01221866657336553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,128,0.028560890091790095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,64,0.012880889077981314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,10240,128,0.0332568883895874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,10240,32,0.01348888874053955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,64,0.02832533253563775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,32,0.028562668297025893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,10240,256,0.030596445004145306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,65536,1.1238817638821073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,51200,0.8559261957804362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,65536,2.5820541381835938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,65536,2.481920030381944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,51200,2.0520267486572266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,16384,0.30665866533915204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,16384,0.5949004491170248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,12288,0.2286613252427843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,16384,0.639901320139567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,51200,1.9440996381971571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,12288,0.4392880068884955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,12288,0.4862382147047255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,10240,0.20885333749983045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,8192,0.1670088900460137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,10240,0.3639884524875217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,7168,0.1376933389239841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,10240,0.4087973435719808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,8192,0.29216355747646755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,7168,0.25122666358947754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,6144,0.11927999390496148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,7168,0.2866284317440457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,5120,0.10139289167192246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,6144,0.21335466702779135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,6144,0.24570756488376191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,4096,0.081295112768809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,5120,0.17703644434611002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,5120,0.2069466643863254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,3584,0.07078933053546481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,4096,0.17094843917422822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,4096,0.14309688409169516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,8192,0.3267893261379666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,3584,0.12778045071495905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,3072,0.06075466341442532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,3584,0.15316532717810735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,2560,0.05085777905252245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,3072,0.11239111423492432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,3072,0.13515644603305393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,2048,0.04206933246718513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,2560,0.09707911147011651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,1536,0.033428443802727595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,2560,0.1174515618218316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,2048,0.08087111181683011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,2048,0.09996977779600356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,1024,0.02370311154259576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,1536,0.06373333268695407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,768,0.020199111766285367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,1024,0.06440533532036676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,1024,0.04717155628734165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,1536,0.08308266931109957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,512,0.01512888901763492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,768,0.03872266742918227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,768,0.05536444319619072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,256,0.011811555259757571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,512,0.046273777882258095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,128,0.010062221851613786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,512,0.03132355544302199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,256,0.026622222529517278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,256,0.036834667126337685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,64,0.009330666727489894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,8192,32,0.00941955546538035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,128,0.025049777494536504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,8192,128,0.03057510985268487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,32,0.02479022244612376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,65536,1.0831111272176106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,8192,64,0.025055999557177227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,51200,0.8702088991800944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,65536,2.372555626763238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,16384,0.27831554412841797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,51200,1.889510260687934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,65536,2.3487430148654513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,12288,0.21873333719041613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,16384,0.5500373310512966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,12288,0.40653689702351886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,16384,0.6074737972683376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,10240,0.17732532819112143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,12288,0.46063465542263454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,51200,1.8463412390814886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,8192,0.15350488821665445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,10240,0.3397840129004584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,10240,0.38795823521084255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,8192,0.2719377676645915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,7168,0.12216888533698188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,8192,0.30945245424906415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,6144,0.11042666435241699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,7168,0.23330844773186576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,7168,0.2713822258843316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,6144,0.19815556208292642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,5120,0.09011822276645237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,6144,0.23276355531480578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,4096,0.07589244180255465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,5120,0.16487822267744276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,5120,0.19644622008005777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,4096,0.13278755876753065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,3584,0.06358933448791504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,4096,0.16218666235605875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,3072,0.05500266618198819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,3584,0.11791022618611653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,3584,0.1451102230283949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,3072,0.10502844386630589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,2560,0.04669777883423699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,3072,0.1284444464577569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,2048,0.039177778694364764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,2560,0.09048444694942898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,2560,0.11173333062065972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,1536,0.027865777413050335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,2048,0.07499644491407606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,2048,0.09513066874610053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,1024,0.020376889242066275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,1536,0.07827999856736925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,1536,0.059120886855655246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,1024,0.04421155651410421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,1024,0.061089780595567494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,768,0.017966222431924608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,512,0.014946666028764514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,768,0.03638666537072923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,768,0.053096890449523926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,256,0.0120666664507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,512,0.04448355568779839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,512,0.029297779003779095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,256,0.02478222217824724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,128,0.010753778119881948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,256,0.035529779063330755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,64,0.009317333499590555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,128,0.023528888821601868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,7168,128,0.030516445636749268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,7168,32,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,64,0.02332533399264018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,7168,32,0.023393778337372675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,65536,0.8774835798475477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,51200,0.695634682973226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,65536,2.182964536878798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,65536,2.2197227478027344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,51200,1.7259741889105902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,16384,0.23585687743292916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,12288,0.1658880048327976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,16384,0.509444448682997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,16384,0.5743057992723253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,12288,0.37781511412726504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,10240,0.15298577149709067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,51200,1.7413075764973958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,8192,0.12079732947879368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,10240,0.36742933591206867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,10240,0.3149226771460639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,7168,0.10537955496046279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,8192,0.24814844131469727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,8192,0.29359110196431476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,6144,0.09271111090977986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,7168,0.21645332707299125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,7168,0.2558355596330431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,5120,0.07559911410013835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,6144,0.18325066566467285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,12288,0.4362364345126682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,4096,0.0632551113764445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,6144,0.22122044033474395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,5120,0.15221510993109807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,3584,0.05356711149215698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,5120,0.18573244412740073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,4096,0.12287022007836236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,4096,0.15362311734093562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,3072,0.045828445090187915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,3584,0.10898400015301174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,3584,0.13722488615247938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,2560,0.038479109605153404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,3072,0.1213644478056166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,2048,0.032989333073298134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,3072,0.09585333532757229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,2560,0.08342488606770833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,1536,0.025288888149791296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,2048,0.06916977961858113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,2048,0.09115999937057495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,1536,0.054733332660463124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,1024,0.018981332580248516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,1536,0.07462577687369452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,768,0.016314667132165697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,1024,0.039832890033721924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,1024,0.058973332246144615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,768,0.03387466735310025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,512,0.011613333390818702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,768,0.05111022127999199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,512,0.026501332720120747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,512,0.04289422101444668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,2560,0.10522311263614231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,256,0.034332444270451866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,256,0.02388266722361247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,128,0.008344888687133789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,6144,128,0.028776887390348647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,128,0.022307554880777996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,64,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,32,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,64,0.022037333912319605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,6144,32,0.02164266672399309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,6144,256,0.00906133320596483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,65536,0.7454231050279406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,51200,0.5891911188761393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,65536,1.9727484385172527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,65536,2.092794630262587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,51200,1.5513742234971788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,16384,0.213919997215271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,16384,0.4641395674811469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,16384,0.5406524340311686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,12288,0.14590488539801702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,51200,1.6391368442111547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,10240,0.13896355364057753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,12288,0.34652090072631836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,12288,0.4119857682122125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,10240,0.28884532716539174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,8192,0.11098488834169175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,10240,0.3465422259436713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,7168,0.09818844662772284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,8192,0.22884888119167754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,8192,0.27633688184950084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,7168,0.19867555300394693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,6144,0.08335733413696289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,7168,0.24127644962734648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,6144,0.16766577296786836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,5120,0.07051289081573486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,6144,0.2080391115612454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,4096,0.05590933561325073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,5120,0.17484800020853677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,3584,0.047826667626698814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,4096,0.11362577809227838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,4096,0.14567822880215114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,3584,0.10027733114030625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,3072,0.042437334855397545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,3072,0.08790666527218288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,3584,0.1304533349143134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,2560,0.03453600075509813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,3072,0.11464444796244304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,2048,0.029763556189007227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,2560,0.07404977745480008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,2560,0.09920889139175415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,1536,0.023210666245884363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,2048,0.06318310896555583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,5120,0.1400417751736111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,1024,0.01685066686736213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,2048,0.08601155545976426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,1536,0.049792889091703624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,768,0.013906665974193148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,1536,0.07067555851406522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,1024,0.03690933187802633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,1024,0.05615022447374132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,512,0.010781333678298526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,768,0.031160887744691636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,768,0.0483457777235243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,256,0.008277333445019191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,512,0.03872266742918227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,512,0.025263999899228413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,128,0.007387555307812161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,256,0.03295200069745382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,256,0.020977778567208186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,64,0.006602666858169768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,5120,32,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,5120,128,0.0273973329199685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,128,0.020289777053727042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,64,0.020069332586394418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,5120,32,0.019892444213231403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,65536,0.5965341991848415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,51200,0.437682655122545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,65536,1.7667377259996202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,51200,1.3967119852701824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,16384,0.1500355535083347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,65536,1.9611795213487413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,51200,1.5375120374891493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,16384,0.42449333932664657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,12288,0.12098311053382026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,10240,0.10323644346661037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,16384,0.5101822217305502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,12288,0.3159688843621148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,12288,0.3874444431728787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,8192,0.07909422450595431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,10240,0.2637573348151313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,7168,0.07124000125461154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,10240,0.3270302348666721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,8192,0.20897244082556832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,8192,0.2595493263668484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,6144,0.061864886018964976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,7168,0.18073422378963896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,7168,0.2275528907775879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,5120,0.05468533436457316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,6144,0.15350843800438776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,6144,0.1953271097607083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,4096,0.04337333308325874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,5120,0.16424800290001765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,5120,0.12856354978349474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,3584,0.03764355513784621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,4096,0.10401866833368938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,4096,0.13613066408369276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,3072,0.03281155559751723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,3584,0.12233332792917888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,3584,0.09283110830518936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,2560,0.027245332797368366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,3072,0.08111644453472562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,3072,0.10767200258043076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,2048,0.023760888311598036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,2560,0.0932391087214152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,2560,0.06881511211395264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,2048,0.05521066652403938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,1536,0.01869155632125007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,2048,0.08043289184570312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,1024,0.014318222800890604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,1536,0.0452186663945516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,1536,0.06682044267654419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,768,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,1024,0.033702221181657575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,1024,0.0507262216673957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,512,0.009132444030708736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,768,0.04545155498716566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,768,0.02918755677011278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,256,0.007037333316273159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,512,0.02300888962215847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,512,0.03669688767857022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,128,0.006287999865081575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,256,0.03122399912940131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,256,0.019662222928471036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,64,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,128,0.018623110320832994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,4096,32,0.0059004442559348206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,4096,128,0.027218666341569688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,64,0.018605333235528734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,65536,0.5379075474209255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,4096,32,0.01828266680240631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,51200,0.42784444491068524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,65536,1.6663458082411025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,65536,1.9604487948947484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,16384,0.14210311571756998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,51200,1.307700475056966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,16384,0.4033679962158203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,12288,0.10695466730329727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,51200,1.5388844807942708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,16384,0.5078915490044488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,10240,0.09245955281787449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,12288,0.3016035556793213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,8192,0.07260177532831828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,12288,0.38754577106899685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,10240,0.2514622211456299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,10240,0.32590844896104604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,7168,0.06325066751903959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,8192,0.19945066505008271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,8192,0.2600239912668864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,6144,0.05516977773772346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,7168,0.17304621802435982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,7168,0.2262079980638292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,5120,0.04758044415050083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,6144,0.1466968854268392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,6144,0.1949875619676378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,4096,0.038903110557132296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,5120,0.12189243899451362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,3584,0.03468622101677789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,5120,0.16443467140197754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,4096,0.09910311301549275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,3072,0.03186310993300544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,3584,0.08821688758002387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,3584,0.1217155588997735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,2560,0.02695466743575202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,4096,0.13644356197781032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,3072,0.07683999670876397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,2048,0.023009777069091797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,3072,0.10744711425569321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,2560,0.06479555368423462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,1536,0.018026666508780587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,2048,0.07961955335405138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,2048,0.05246222350332472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,1024,0.013912889692518445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,1536,0.043065776427586876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,1536,0.06572888957129584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,768,0.011023110813564725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,1024,0.03229422370592753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,1024,0.051853332254621715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,512,0.009007111191749573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,768,0.027075555589463975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,768,0.04344444473584493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,256,0.007351111206743453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,512,0.02200444373819563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,512,0.03631822268168131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,256,0.018554665976100497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,256,0.02997066577275594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,128,0.0063573333124319715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,128,0.017839110559887357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,2560,0.09354488717185126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,64,0.00554044461912579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3584,32,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,32,0.017490666773584154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3584,64,0.017980444762441847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,65536,0.4259804354773627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3584,128,0.025835555460717943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,51200,0.35543733172946507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,65536,1.5634915033976238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,51200,1.2420408460828993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,65536,1.831226560804579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,16384,0.11662222279442681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,16384,0.38099198871188694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,12288,0.08312444554434882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,16384,0.4774746894836426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,51200,1.4381173451741536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,10240,0.08341511090596516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,12288,0.2871368991004096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,12288,0.3630666732788086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,8192,0.06369244390063815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,10240,0.3056844340430366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,10240,0.2406568792131212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,7168,0.057077334986792676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,8192,0.1904533306757609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,8192,0.24288622538248697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,6144,0.045906666252348155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,7168,0.21294221613142225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,7168,0.16550310452779135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,5120,0.03860622313287523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,6144,0.13981689347161189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,6144,0.1822364462746514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,4096,0.03244266576237149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,5120,0.15436532762315539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,5120,0.11670045057932536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,3584,0.029185778564876978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,4096,0.09438400136099921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,4096,0.1273182233174642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,3072,0.02451466686195797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,3584,0.11477066410912408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,3584,0.08409066994984944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,3072,0.07290310992134942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,2560,0.02171733313136631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,3072,0.10092000166575114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,2048,0.01852799952030182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,2560,0.06311377551820543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,2048,0.051332443952560425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,2560,0.08824177583058675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,1536,0.014583999911944071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,2048,0.07588711049821642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,1024,0.01075111081202825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,1024,0.03052622079849243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,1536,0.04001422060860528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,1024,0.047945777575174965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,768,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,1536,0.06122666597366333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,512,0.007319110963079665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,768,0.02570755614174737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,768,0.04119644562403361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,256,0.006264888991912206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,512,0.035791112316979304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,128,0.005555555638339784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,512,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,256,0.01755555636352963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,256,0.029892444610595703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,64,0.0052533331844541765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,3072,32,0.0052666668262746595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,3072,128,0.025360888904995386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,64,0.016490666402710807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,32,0.016507556041081745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,65536,0.39447911580403644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,51200,0.29539288414849174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,3072,128,0.016893333858913846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,65536,1.4734924104478624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,16384,0.10156266556845771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,51200,1.1482488844129775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,16384,0.36206756697760684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,65536,1.8307821485731337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,12288,0.07453244262271456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,12288,0.2726817660861545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,12288,0.36232177416483563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,10240,0.07084000110626221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,16384,0.47730668385823566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,8192,0.05839911434385511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,10240,0.22753511534796822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,10240,0.3059768941667345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,7168,0.049511998891830444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,8192,0.18051466676923963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,8192,0.24375645319620767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,6144,0.043346666627460055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,7168,0.1566151115629408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,6144,0.13225777943929037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,7168,0.21269244617886016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,5120,0.036631110641691424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,6144,0.1823528872595893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,51200,1.4357564714219835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,4096,0.02923822071817186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,5120,0.15332799487643772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,4096,0.08942755725648667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,3584,0.025164445241292317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,5120,0.11044533385170831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,4096,0.1270293394724528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,3072,0.02351288828584883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,3584,0.07991911305321588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,3584,0.1141128937403361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,2560,0.019516444868511625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,3072,0.07028977738486396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,3072,0.10054222080442642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,2048,0.016153777639071148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,2560,0.08732444710201687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,2560,0.05973777506086561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,1536,0.01297155519326528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,2048,0.0752568907207913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,1536,0.06035733222961426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,1536,0.03822755482461717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,1024,0.009454222189055549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,1024,0.04741600155830383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,768,0.008101333346631791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,768,0.024478221933046978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,2048,0.047818667358822294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,512,0.006894222150246303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,768,0.04080533319049411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,512,0.019967110620604623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,512,0.03453155689769321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,256,0.005618666609128316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,256,0.01657777859105004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,1024,0.028969777954949275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,128,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,128,0.024673778149816725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,64,0.004970666848950916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,128,0.016229333149062265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2560,32,0.004924444274769889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,64,0.015582222077581616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2560,32,0.016114667057991028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,65536,0.29492444462246364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2560,256,0.028900444507598877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,51200,0.24227110544840494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,65536,1.3493821885850694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,51200,1.0576391220092773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,16384,0.08333066436979505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,65536,1.702407095167372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,16384,0.3408506711324056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,12288,0.06403111086951362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,16384,0.44483200709025067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,12288,0.2577582200368245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,51200,1.3375235663519964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,10240,0.04982133375273811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,10240,0.28523556391398114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,10240,0.21607822842068145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,8192,0.17051733864678276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,7168,0.035952889257007174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,8192,0.2283973428938124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,12288,0.338318215476142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,6144,0.0326275560590956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,7168,0.14765066570705837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,7168,0.1986008882522583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,6144,0.12500977516174316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,5120,0.027851555082533095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,8192,0.042865776353412204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,4096,0.022646221849653456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,5120,0.1036284433470832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,6144,0.1701271136601766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,5120,0.14302399423387316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,3584,0.021374222305085924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,4096,0.08424444331063165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,3072,0.017832888497246634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,4096,0.11839644114176433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,3584,0.07468622260623507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,2560,0.01611199975013733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,3072,0.06531111399332683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,3584,0.10761421918869019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,3072,0.09457688861423069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,2048,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,2560,0.05451200074619717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,1536,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,2560,0.08208355638715956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,2048,0.04456355505519443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,1024,0.008669333325492011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,2048,0.07169600327809651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,1536,0.03500888744990031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,1536,0.05784177780151367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,768,0.007988444632954067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,1024,0.02681422233581543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,512,0.0062773335311147906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,768,0.0229777776532703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,1024,0.04676622152328491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,256,0.0052560000783867305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,512,0.018924444913864136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,512,0.03291911217901442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,128,0.004528000122971005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,256,0.027756444282001917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,128,0.015184000134468079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,256,0.01611199975013733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,128,0.024352888266245525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,64,0.0041857775714662345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,2048,32,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,32,0.01484444406297472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,65536,0.244758235083686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,2048,768,0.038747555679745145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,2048,64,0.015242666006088257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,65536,1.248684459262424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,51200,0.18326932854122588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,65536,1.7011502583821614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,51200,0.9800444708930122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,16384,0.06706488794750638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,16384,0.3201199902428521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,12288,0.04539466566509671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,16384,0.44457419713338214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,51200,1.3358951144748266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,10240,0.04148711098564996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,12288,0.2427368958791097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,12288,0.3384231196509467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,8192,0.03285333183076646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,10240,0.20356355773078072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,10240,0.2846604453192817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,7168,0.029554665088653564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,8192,0.16051733493804932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,8192,0.22713867823282877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,7168,0.13880266083611384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,6144,0.02477955487039354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,7168,0.19794400533040366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,5120,0.02196177840232849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,6144,0.11712888876597087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,6144,0.1702124410205417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,5120,0.09716088904274835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,4096,0.018439999885029264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,5120,0.142795549498664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,3584,0.01681511104106903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,4096,0.0789804458618164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,4096,0.11850044462415908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,3584,0.0699439975950453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,3072,0.014679110712475248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,3584,0.10621333122253418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,2560,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,3072,0.060757332377963595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,3072,0.09556355741288926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,2560,0.05106755428844028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,2048,0.010614222122563256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,2560,0.08154044548670451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,1536,0.009112888740168678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,2048,0.041915555795033775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,2048,0.07087822092903985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,1536,0.03300889001952277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,1024,0.007463110817803278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,1536,0.05829066700405545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,768,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,1024,0.025454221500290766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,1024,0.04460444384151035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,512,0.005629333357016246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,512,0.01796533332930671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,512,0.03301422132386102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,256,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,256,0.015284443895022074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,768,0.0380711125002967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,128,0.004327110946178436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,256,0.028047998746236164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,128,0.01422133379512363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,64,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1536,32,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1536,128,0.023805333508385554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,64,0.014228444960382251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,32,0.013842666314707862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,65536,0.1696488857269287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,51200,0.13688266277313232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1536,768,0.021653332644038733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,65536,1.1616346571180556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,16384,0.048579557074440845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,51200,0.9129440519544813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,65536,1.5718293719821508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,16384,0.3009813361697727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,12288,0.033369776275422834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,51200,1.2347830666436088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,16384,0.41230045424567324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,10240,0.029141333368089464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,12288,0.22757066620720756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,8192,0.022224000758594934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,12288,0.31342045466105145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,10240,0.2645244333479139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,7168,0.02047288914521535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,10240,0.19112622737884521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,8192,0.15020267168680826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,8192,0.2107084459728665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,6144,0.017284444636768766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,7168,0.1295128928290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,5120,0.015454222758611044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,7168,0.1832808918423123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,6144,0.10952799850040013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,6144,0.1566586626900567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,4096,0.013405332962671915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,5120,0.09052622318267822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,3584,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,5120,0.1323795583513048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,4096,0.07354311148325603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,4096,0.10971733596589829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,3072,0.010586666564146677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,3584,0.06540622313817342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,2560,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,3584,0.09855733315149943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,3072,0.0872337751918369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,2560,0.0752426650789049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,2048,0.007968000239796108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,2560,0.04771555463473002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,2048,0.038776887787712946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,1536,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,2048,0.06557599703470866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,1536,0.030844443374209937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,1024,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,1536,0.053527110152774386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,3072,0.05621777640448677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,1024,0.023797333240509033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,768,0.0053093334039052325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,1024,0.04166400101449754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,512,0.004908444566859139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,768,0.02067466742462582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,768,0.03605955508020189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,256,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,512,0.031215998861524794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,256,0.014273777604103088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,512,0.01702577703528934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,128,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,256,0.026447110705905493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,64,0.003584889074166616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,1024,128,0.02231377859910329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,1024,32,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,64,0.01346666696998808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,32,0.01311644415060679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,1024,128,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,65536,1.1220799552069771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,65536,0.14058666759067112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,51200,0.11068799760606553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,51200,0.8818773163689507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,16384,0.0384808878103892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,51200,1.2319893307156033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,65536,1.5718924204508464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,16384,0.2905742327372233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,12288,0.029330667522218492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,12288,0.22023110919528535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,16384,0.4105902247958713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,10240,0.022728888524903193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,12288,0.31308089362250435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,8192,0.018999111321237352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,10240,0.1842471096250746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,10240,0.2639182143741184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,7168,0.017231111725171406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,8192,0.14558488792843288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,8192,0.2101057767868042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,6144,0.01462933255566491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,7168,0.12562666998969182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,6144,0.10565688874986436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,7168,0.1824702156914605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,6144,0.15615110927157932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,5120,0.013033777475357056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,4096,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,5120,0.1314880053202311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,5120,0.08789155880610149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,3584,0.010057777994208867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,4096,0.10962311426798503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,3584,0.06293866369459364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,3584,0.09722044732835557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,3072,0.009096889032257928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,3072,0.05424000157250298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,2560,0.008047999607192146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,3072,0.08727644549475776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,2048,0.007101332975758447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,2560,0.07536088758044772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,2560,0.04488799969355265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,1536,0.006671999891599019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,2048,0.06358310911390516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,2048,0.03764266769091288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,1536,0.029367112451129492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,1024,0.0053288886944452924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,1536,0.05231022172504001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,768,0.004917333523432414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,1024,0.022872888379626807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,1024,0.04158488909403483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,768,0.0199297782447603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,512,0.004350222233268949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,768,0.03575377662976583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,256,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,512,0.016566221912701923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,512,0.030801776382658217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,128,0.00360622215602133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,256,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,256,0.0262435558769438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,64,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,4096,0.0708728896247016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,768,128,0.022443556123309668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,768,32,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,128,0.01311377767059538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,64,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,65536,0.12059733602735732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,768,32,0.012709333664841123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,51200,0.09649866819381714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,51200,0.8469200134277344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,16384,0.03279022375742594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,65536,1.0820070902506511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,51200,1.2228711446126301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,16384,0.2804204357994927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,12288,0.02402666707833608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,65536,1.5582008361816406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,16384,0.4087306658426921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,10240,0.019951999187469482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,12288,0.21256532933976915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,12288,0.31047556135389537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,8192,0.017217778497272067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,10240,0.1780008872350057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,10240,0.2623448901706272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,7168,0.01500888831085629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,8192,0.14060533046722412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,8192,0.20884533723195395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,7168,0.121035549375746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,7168,0.18096889389885795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,6144,0.10136889086829291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,6144,0.15568621953328451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,5120,0.011687111523416309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,5120,0.08406933148701985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,4096,0.009676444033781687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,4096,0.06829066409005059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,6144,0.013175999952687157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,4096,0.10915288660261367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,3584,0.00907022257645925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,3584,0.060180445512135826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,3072,0.007972444097201029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,3584,0.097871998945872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,5120,0.13142311573028564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,3072,0.05146222313245138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,2560,0.007663110891977946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,3072,0.08630933364232381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,2048,0.006667555620272954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,2560,0.04348088966475593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,2560,0.07477777534061007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,1536,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,2048,0.06347733073764376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,2048,0.035291555855009295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,1536,0.028354667954974707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,1024,0.005010666532648934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,1536,0.05151289039187961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,768,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,1024,0.02204533252451155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,1024,0.04122399952676561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,512,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,768,0.01924888955222236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,768,0.03567022085189819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,256,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,512,0.030605332718955144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,512,0.016182222300105624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,256,0.01316177762216992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,128,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,256,0.025802666942278545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,64,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,128,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,512,32,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,512,128,0.022112000319692824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,64,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,65536,0.0673440032535129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,512,32,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,51200,0.05406844284799364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,51200,0.8151946597629123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,65536,1.038779576619466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,16384,0.0218551109234492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,16384,0.2699768808152941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,65536,1.5596257315741644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,12288,0.017603556315104168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,16384,0.4071555667453342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,12288,0.20509778128729927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,51200,1.22325865427653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,10240,0.015028445257080926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,12288,0.31024622917175293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,8192,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,10240,0.17162044843037924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,7168,0.011325333681371478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,10240,0.26145243644714355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,8192,0.20863021744622123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,8192,0.13588621881273058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,7168,0.11645778020222981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,6144,0.010216000179449717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,5120,0.009032888544930352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,6144,0.09786844253540039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,6144,0.15529244475894505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,4096,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,5120,0.08081066608428955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,5120,0.13041067123413086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,3584,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,4096,0.06565422481960721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,4096,0.10774399836858113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,3072,0.006912888752089606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,3584,0.05782933367623223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,3584,0.09699200259314643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,3072,0.04892888996336195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,2560,0.00639111093348927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,3072,0.08519199821684097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,2048,0.005468444277842839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,2560,0.040532443258497454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,2560,0.07406666543748644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,1536,0.0052622221410274506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,2048,0.03394577900568644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,2048,0.06343111064698961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,7168,0.18092711766560873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,1536,0.027404444085227117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,1024,0.004269333349333869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,1536,0.05135644475618998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,768,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,1024,0.021542222963439092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,1024,0.04085866610209147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,768,0.018567999203999836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,768,0.0350835555129581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,512,0.01540177729394701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,512,0.02991466720898946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,256,0.025125333004527624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,128,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,128,0.012087999946541257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,256,128,0.021371554997232225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,64,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,512,0.003980444537268745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,256,32,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,64,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,65536,0.05552355448404948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,32,0.01181777815024058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,256,256,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,51200,0.044907556639777295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,65536,1.0384622149997287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,51200,0.814068423377143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,16384,0.013227555486891003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,16384,0.26965422100490993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,65536,1.5596195856730144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,12288,0.013184888495339287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,16384,0.40704621209038633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,51200,1.2222461700439453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,12288,0.20495733949873182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,10240,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,12288,0.3103128804100884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,8192,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,10240,0.26137423515319824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,10240,0.1715679963429769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,7168,0.0111191107167138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,8192,0.13481067286597356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,8192,0.20738666587405732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,6144,0.010006222460005019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,7168,0.11591822571224636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,7168,0.1810631089740329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,5120,0.00906755526860555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,6144,0.09689688682556152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,6144,0.15525600645277235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,5120,0.08030311266581218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,5120,0.13004177146487766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,4096,0.007692444655630324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,3584,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,4096,0.06514755884806316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,4096,0.10709155268139309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,3072,0.006614222294754452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,3584,0.05774133072959053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,3584,0.0964319970872667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,2560,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,3072,0.048525333404541016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,3072,0.08558133575651382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,2560,0.07147288984722562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,2560,0.03943199912707011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,2048,0.005256888767083486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,1536,0.00461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,2048,0.033279998434914485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,2048,0.061627553568945996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,1024,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,1536,0.027450667487250432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,1536,0.05121066504054599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,768,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,1024,0.02107911143038008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,1024,0.04045155644416809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,768,0.03509066502253214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,512,0.0036159998012913596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,512,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,256,0.0033404444240861465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,512,0.029969778325822618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,128,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,256,0.024700444605615404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,256,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2304,128,128,0.02166755497455597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,128,32,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,64,0.011720889144473605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,65536,0.054281777805752225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,32,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,51200,0.04332977864477369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,768,0.018560000591807894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,16384,0.014507555299335055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,16384,0.2695964442359077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,12288,0.012059555285506778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,128,128,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,51200,0.8136809137132434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,10240,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,65536,1.0376880433824327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,12288,0.2044515609741211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,8192,0.00977244476477305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,7168,0.009340444372759925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,6144,0.00867911097076204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,10240,0.17075910833146837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,8192,0.13504710462358263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,5120,0.008746667040718926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,6144,0.09665866692860921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,7168,0.1157075564066569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,4096,0.007332444190979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,3584,0.0070159998204973005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,5120,0.08035377661387126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,3072,0.006421333385838403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,4096,0.06538399722841051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,3584,0.05755111243989733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,3072,0.048231111632453076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,2560,0.00563822231358952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,2048,0.005295111073387994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,1536,0.004610666798220741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,2560,0.04081066780620151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,1024,0.003923555629120933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,1536,0.027249778310457867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,768,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,1024,0.021298666795094807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,512,0.0032559999575217566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,768,0.018290667070282828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,256,0.0032586666444937387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,512,0.015223999818166098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,128,0.002962666667169995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,256,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,64,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,128,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,64,32,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,64,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,65536,0.05418577790260315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,2048,0.03348888953526815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,64,32,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,51200,0.04078755444950528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,16384,0.010066666536860997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,16384,0.26963554488288033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,51200,0.813470204671224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,65536,1.0381351047092013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,12288,0.008803555534945594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,10240,0.007999111380841997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,8192,0.012851555314328937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,12288,0.20437688297695586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,7168,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,8192,0.13467110527886286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,6144,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,10240,0.17137066523234049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,5120,0.010382222632567087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,7168,0.11598222785525852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,4096,0.009405333134863112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,6144,0.09686844216452704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,3584,0.008671999805503422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,5120,0.08058844672309028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,4096,0.06462755468156603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,3072,0.007616000043021307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,3584,0.05700088871849907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,2560,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,3072,0.048398223188188344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,2560,0.03942222065395779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,1536,0.004670222186379963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,2048,0.03329688972897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,1536,0.027122666438420612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,1024,0.003966222206751506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,768,0.003671110918124517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,1024,0.021270222134060327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,512,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,768,0.01869066721863217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,256,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,2048,0.004907555464241239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,128,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,256,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,64,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,128,0.011855999628702799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,64,0.01127822200457255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2304,32,32,0.0029688889367712867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,32,0.011491555306646558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2304,32,512,0.015180443723996481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,16384,2.0870880550808377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,12288,1.5384755664401586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,16384,3.002456876966688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,10240,1.2961066563924153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,51200,6.454068501790364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,12288,2.080866707695855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,8192,1.104582256740994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,10240,1.7101075914171007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,8192,1.3328559663560655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,7168,0.9275413089328342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,6144,0.7863093482123481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,7168,1.1481866836547852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,5120,0.6643688943650987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,4096,0.5367057588365343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,5120,0.813476456536187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,51200,10.268974304199219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,3584,0.4737146695454915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,4096,0.6510728730095757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,3072,0.43165866533915204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,3584,0.567829344007704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,6144,0.9791368908352323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,3072,0.48998133341471356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,2560,0.36155022515190977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,2048,0.2995937665303548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,1536,0.22816620932685006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,2048,0.3410328759087457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,2560,0.4165991147359212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,1024,0.1595324410332574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,768,0.1189182202021281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,1536,0.2705235481262207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,1024,0.20447644922468397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,512,0.09270577960544163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,256,0.06960088676876493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,768,0.16437421904669866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,128,0.05466755562358432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,512,0.13972711563110352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,64,0.047332445780436196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,128,0.11872800191243489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,256,0.12493956089019775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,65536,32,0.049323555496003896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,64,0.1201368835237291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,65536,32,0.1216942204369439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,51200,5.0330424838595915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,65536,6.505296071370442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,16384,1.6982488632202148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,12288,1.268964449564616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,16384,2.361999935574002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,10240,1.10535822974311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,65536,10.435866461859808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,12288,1.6619466145833333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,8192,0.898903104994032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,10240,1.3569777806599934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,7168,0.7953991360134549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,8192,1.0542204115125868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,7168,0.918916490342882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,6144,0.591933356391059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,6144,0.7817760043674045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,5120,0.47307289971245664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,51200,8.112498813205296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,4096,0.39509601063198513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,5120,0.6568622059292263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,3584,0.3699599901835124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,4096,0.5337031152513292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,3072,0.30421778890821666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,2560,0.24353599548339844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,3584,0.4556444485982259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,2048,0.19877333111233184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,3072,0.3944648901621501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,2048,0.27431021796332467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,1536,0.15406755606333414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,1024,0.10802844497892593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,768,0.08641866842905681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,1536,0.21816444396972656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,1024,0.1658044391208225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,512,0.06630311409632365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,768,0.13279910882314047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,256,0.05074310965008206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,128,0.04392533169852363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,512,0.11243732770284016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,64,0.03745244608985053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,256,0.10092800193362766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,51200,32,0.039474666118621826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,128,0.0961315565639072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,64,0.09701155291663276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,2560,0.35942755805121523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,51200,32,0.09775022003385757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,51200,1.6268604066636827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,65536,2.126911163330078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,16384,0.5497182210286459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,12288,0.396213346057468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,16384,0.8743084271748861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,51200,3.0092976888020835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,10240,0.438918219672309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,10240,0.521017763349745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,65536,3.902709537082248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,8192,0.3489893277486165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,7168,0.3206000063154432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,8192,0.4092675579918756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,7168,0.3574479950798883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,6144,0.2652124563852946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,5120,0.21451821592119005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,5120,0.25442666477627224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,4096,0.17679822444915771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,12288,0.6303564707438151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,3584,0.1564799944559733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,4096,0.204767107963562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,3072,0.13304800457424587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,2560,0.10188088814417522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,3584,0.1805582178963555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,2048,0.06999288664923774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,3072,0.15767378277248806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,1536,0.057576000690460205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,2560,0.13628799385494655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,6144,0.3057520124647352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,2048,0.11260799566904704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,1024,0.042134222057130605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,768,0.03292799989382426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,512,0.024894222617149353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,1024,0.06857777966393365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,256,0.019291554888089497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,768,0.05698577562967936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,128,0.01833599971400367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,256,0.040693332751592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,512,0.047078222036361694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,64,0.015291555060280694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,16384,32,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,128,0.038153777519861855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,64,0.037837333149380155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,32,0.03694577680693732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,16384,1536,0.09003644519382054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,65536,1.5789563920762804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,51200,1.2393582661946614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,16384,0.39143021901448566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,65536,3.084636476304796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,12288,0.317230224609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,16384,0.7031262185838488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,51200,2.4361555311414933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,12288,0.5122195349799262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,10240,0.23670665423075357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,8192,0.19712800449795195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,10240,0.4249422285291884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,8192,0.3333253330654568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,7168,0.19096355968051484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,6144,0.16765688525305855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,7168,0.29106489817301434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,5120,0.13903289371066624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,6144,0.24938755565219453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,4096,0.11245689127180312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,3584,0.08840888738632202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,5120,0.20892088943057588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,4096,0.16915555795033774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,3584,0.1489644447962443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,2560,0.06453955835766263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,2560,0.11260710822211371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,3072,0.13046844800313315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,2048,0.05314844515588549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,1536,0.04162044326464335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,2048,0.09360533290439182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,1536,0.07494133048587374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,1024,0.030850665436850652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,768,0.02381422287887997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,1024,0.055919110774993896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,3072,0.07614311244752672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,512,0.018975999620225694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,256,0.014761777387724983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,768,0.0466284453868866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,128,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,256,0.032390223609076604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,512,0.03716533382733663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,64,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,12288,32,0.014870221416155497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,128,0.030344890223609075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,64,0.030023998684353296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,12288,32,0.029529776838090684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,51200,1.0612302356296115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,65536,1.3317546844482422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,16384,0.3521537780761719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,12288,0.2629893355899387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,16384,0.6186809009975857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,51200,2.130964491102431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,10240,0.22729688220553926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,12288,0.4550471305847168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,8192,0.1833475563261244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,65536,2.7002391815185547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,10240,0.3761288854810927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,7168,0.16049688392215306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,6144,0.13879111078050402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,8192,0.29602400461832684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,5120,0.11632088820139568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,7168,0.2588702307807075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,4096,0.09397066964043511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,6144,0.22157955169677734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,3584,0.08373510837554932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,5120,0.18467644850413004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,3072,0.07292622327804565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,4096,0.14980444643232557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,2560,0.06262755393981934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,2048,0.05127822359402975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,3584,0.13347555531395808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,2560,0.10051200124952529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,3072,0.1165448824564616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,2048,0.08397510978910659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,1024,0.02534577747186025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,768,0.02124533388349745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,1536,0.0666560000843472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,512,0.01573600040541755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,1024,0.04981244603792826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,256,0.01292711082432005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,768,0.04085066583421495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,512,0.03320711188846164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,128,0.010824888944625854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,256,0.028202666176689997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,64,0.009497777455382878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,32,0.009711110757456886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,128,0.027120888233184814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,64,0.027494221925735474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,10240,32,0.02682311170630985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,10240,1536,0.03971288932694329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,65536,1.0707297854953342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,51200,0.8188284238179525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,16384,0.2738408777448866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,12288,0.2061448891957601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,16384,0.5611013306511773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,10240,0.17161156071556938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,12288,0.3976764414045546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,51200,1.8260798984103734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,8192,0.14414932992723253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,65536,2.3572105831570096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,7168,0.12917333179050022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,10240,0.3292115529378255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,8192,0.2594177722930908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,6144,0.10702488819758098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,7168,0.2263253264957004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,5120,0.11932178338368733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,6144,0.1927342282401191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,4096,0.09663910998238458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,3584,0.08225778076383802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,5120,0.16147377755906847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,4096,0.1305911143620809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,3072,0.0733591119448344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,2560,0.06252800093756782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,3584,0.11682577927907307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,3072,0.10264088710149129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,2048,0.04158577654096816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,2560,0.0890071127149794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,1536,0.03282577792803446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,1024,0.02442488902144962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,2048,0.07331999805238512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,768,0.018862222631772358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,1536,0.05799377626842923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,512,0.015357333752844067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,1024,0.04384711053636339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,768,0.03674844569630093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,256,0.0114702218108707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,128,0.009737778041097853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,512,0.028873778051800195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,64,0.00870577742656072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,256,0.025256888733969793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,8192,32,0.008669333325492011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,128,0.024022221565246582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,64,0.02399377855989668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,8192,32,0.023708444502618577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,51200,0.7061742146809896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,65536,0.9202577802870008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,16384,0.2486586570739746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,16384,0.5227377679612901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,12288,0.1846639977561103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,65536,2.1745262145996094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,10240,0.15396266513400608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,12288,0.36872177653842503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,51200,1.6808364656236436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,8192,0.12077421612209743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,7168,0.1066604455312093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,10240,0.3063111040327284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,8192,0.24107111824883354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,6144,0.09290399816301133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,5120,0.07864799764421251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,7168,0.20915467209286162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,4096,0.06446133057276408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,5120,0.14987556139628092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,3584,0.05765599674648709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,4096,0.12114666567908393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,3072,0.05011644297175937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,3584,0.10819822549819946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,2560,0.04315200116899279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,3072,0.09580889013078477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,2048,0.03682577941152785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,2560,0.08280799786249797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,1536,0.02935911218325297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,2048,0.0681191086769104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,1024,0.020184000333150227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,1536,0.05430755681461758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,6144,0.17873067326015898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,768,0.015876443849669564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,512,0.013367111484209696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,1024,0.040440890524122454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,256,0.009865778187910715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,768,0.034550223085615374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,128,0.0087333329849773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,512,0.027112000518374976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,256,0.023377777801619634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,64,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,7168,32,0.008388444781303406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,128,0.02236355510022905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,64,0.022288888692855835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,7168,32,0.02197866638501485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,51200,0.6200239923265245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,65536,0.7818711068895129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,16384,0.21467467149098715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,12288,0.1644079950120714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,16384,0.4581262270609538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,10240,0.13894311587015787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,12288,0.3596328894297282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,51200,1.5420257780287001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,8192,0.11842577987247044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,10240,0.28263288074069554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,7168,0.0996088915401035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,6144,0.08591555224524604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,8192,0.22267821100023058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,5120,0.07508266634411283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,65536,1.9577723609076605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,7168,0.19360978073543975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,4096,0.06235111422008938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,6144,0.1649493376413981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,3584,0.05250489049487644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,5120,0.1387422217263116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,3072,0.0455022222465939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,4096,0.1121857828564114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,2560,0.038542220989863075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,3584,0.09986222452587551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,2048,0.031911998987197876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,3072,0.08685866991678874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,2560,0.07662222120496961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,1536,0.02532977859179179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,2048,0.06289066871007283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,1024,0.01900533338387807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,768,0.015782222151756287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,1536,0.049694223536385425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,512,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,1024,0.03720888826582167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,256,0.008634666601816813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,768,0.03259288933542039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,512,0.025388444463411968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,128,0.008059555457697975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,256,0.02162666618824005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,64,0.007287999822033777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,6144,32,0.007375111182530721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,128,0.020970667401949566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,64,0.020672000116772123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,6144,32,0.021007999777793884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,65536,0.6844017770555285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,51200,0.5453439818488227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,16384,0.18228977256351045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,16384,0.42621954282124835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,65536,1.7830310397677953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,12288,0.13428266843159994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,51200,1.3786871168348525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,10240,0.10630577802658081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,8192,0.0858151117960612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,12288,0.31375643942091197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,7168,0.0759048859278361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,10240,0.2601964473724365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,6144,0.06393688917160034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,8192,0.20494488875071207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,7168,0.1771564483642578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,5120,0.0539262228541904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,4096,0.04434488879309761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,6144,0.1517848836051093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,3584,0.0395342210928599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,5120,0.12729422251383463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,3072,0.03459466828240289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,4096,0.10418311092588638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,3584,0.09255111217498779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,2560,0.029486222399605647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,3072,0.08009066846635607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,2560,0.06767111354404025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,1536,0.019350222415394254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,2048,0.05717866950564914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,1024,0.01513777838812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,1536,0.045920888582865395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,768,0.01202311118443807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,1024,0.03460444344414605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,512,0.009416888985368941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,768,0.02918222215440538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,256,0.007618666523032718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,512,0.02345955537425147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,128,0.007045333584149678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,256,0.02028622229894002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,2048,0.024507555696699355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,64,0.006581333362393909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,128,0.019346666004922655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,5120,32,0.006692444284756978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,32,0.018998222218619455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,5120,64,0.019613333874278598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,65536,0.5194240146213108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,51200,0.4148213333553738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,16384,0.1396497752931383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,16384,0.3816995620727539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,65536,1.6080658170912001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,51200,1.2373510996500652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,10240,0.10532710949579875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,12288,0.2858195569780138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,8192,0.08503111203511556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,10240,0.23679823345608184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,7168,0.0740471084912618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,6144,0.06264444192250569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,8192,0.1861626704533895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,7168,0.16169067223866782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,5120,0.05249866512086657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,4096,0.04195644458134969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,6144,0.13780889246198866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,3584,0.03743733300103082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,5120,0.11578310860527886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,12288,0.10274489058388604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,3072,0.032032890452278986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,4096,0.09428177939520942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,2560,0.027100443840026855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,3584,0.08322399854660034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,2048,0.022410665949185688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,1536,0.018175111876593698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,2560,0.06112888786527845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,3072,0.07251377900441487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,1024,0.013719999955760108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,2048,0.05024088753594292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,768,0.010777778095669217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,512,0.008356444537639618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,1536,0.0410515566666921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,1024,0.030939555830425684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,768,0.026649778087933857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,256,0.007038222418891058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,512,0.02167466613981459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,128,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,256,0.018618666463428073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,64,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,128,0.017972444494565327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,4096,32,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,64,0.01759111053413815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,4096,32,0.01758577757411533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,65536,0.4574249055650499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,51200,0.35944265789455837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,16384,0.12699911329481336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,16384,0.3652053409152561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,12288,0.09328355391820271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,65536,1.5071626239352758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,51200,1.17558044857449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,10240,0.08055910799238417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,12288,0.2724408838484022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,8192,0.0625457763671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,7168,0.055148445897632174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,8192,0.17760799990759957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,10240,0.22645955615573457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,6144,0.04762311114205254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,7168,0.15380888515048557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,5120,0.041032890478769936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,4096,0.033624887466430664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,5120,0.11072888639238145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,6144,0.13225332895914713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,3584,0.03053955568207635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,4096,0.09040977557500203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,3072,0.026774222652117412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,2560,0.023582221733199224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,3584,0.08043377929263644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,2048,0.01994400057527754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,2560,0.060119999779595264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,1536,0.016680000556839835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,2048,0.05000799894332886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,1536,0.039862222141689725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,1024,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,768,0.010465777582592435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,1024,0.029487109846538965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,512,0.008676444490750631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,768,0.025847110483381484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,512,0.020861332615216572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,256,0.008000888758235509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,128,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,3072,0.06948266426722209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,256,0.017620444297790527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,64,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3584,32,0.005568888866239124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,128,0.01757422255145179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,64,0.01690933273898231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3584,32,0.016862221890025668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,65536,0.38911999596489805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,51200,0.30611732270982533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,16384,0.10404444403118557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,16384,0.34448356098598903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,12288,0.07723733451631334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,65536,1.4038915634155273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,10240,0.06996711095174153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,12288,0.25899823506673175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,51200,1.1158498128255208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,8192,0.053744001521004565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,7168,0.047255999512142606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,10240,0.21539555655585396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,6144,0.041787554820378624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,8192,0.1696915494071113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,7168,0.146741337246365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,5120,0.03554666704601712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,4096,0.029343999094433252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,6144,0.1251128911972046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,5120,0.10487200154198541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,3584,0.026506667335828144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,3072,0.023826667004161414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,4096,0.08630933364232381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,3584,0.07724799712498982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,3072,0.06703911225001018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,2560,0.020403555697864957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,2048,0.018218666315078735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,1536,0.014677332507239448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,2560,0.056932442718082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,1024,0.01069866700304879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,2048,0.047051555580563016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,1536,0.03724711139996847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,768,0.009663110805882348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,512,0.007385777930418651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,1024,0.02817244331041972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,256,0.006577777779764599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,768,0.024361777636739943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,128,0.005618666609128316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,64,0.00500177757607566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,256,0.01722133325205909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,512,0.020083554916911654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,3072,32,0.004936888813972473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,64,0.016558221644825406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,65536,0.3492400116390652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,51200,0.2775804466671414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,128,0.016359110673268635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,3072,32,0.015818667080667283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,65536,1.3039360046386719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,16384,0.09145155880186294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,12288,0.07029422124226888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,12288,0.2446275552113851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,16384,0.32657867007785374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,10240,0.060715556144714355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,8192,0.04914933443069458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,51200,1.0242658191257052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,7168,0.04379644327693515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,10240,0.20357155799865723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,8192,0.15972888469696045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,6144,0.037576887342664934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,5120,0.0320995549360911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,7168,0.13825599352518717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,4096,0.02695466743575202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,6144,0.11797422832912868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,3584,0.024327110913064744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,5120,0.09930577543046738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,4096,0.08129422532187568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,3072,0.021537777450349595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,3584,0.07214844226837158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,2560,0.01864088906182183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,2048,0.01579466627703773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,2560,0.052954667144351535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,1536,0.012792000340090858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,1536,0.03453866640726725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,1024,0.009721777505344814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,2048,0.0439662237962087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,768,0.008488000267081791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,1024,0.026467555099063452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,512,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,512,0.018901333212852478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,3072,0.06243377923965454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,256,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,128,0.004974222017659081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,768,0.02381333377626207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,64,0.004655111167165968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,256,0.016220443778567843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,128,0.01552088889810774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2560,32,0.004599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,32,0.01516355574131012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,65536,0.31795022222730845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2560,64,0.015915556086434256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,51200,0.25276533762613934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,16384,0.0856142242749532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,16384,0.30631910430060494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,12288,0.06512888934877184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,51200,0.9438835779825846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,10240,0.05509333478079902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,12288,0.23058133655124238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,8192,0.04339111182424757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,8192,0.15019110838572183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,7168,0.03745155533154806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,7168,0.12980088922712538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,6144,0.03106133474244012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,6144,0.1112755537033081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,5120,0.026333333717452154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,5120,0.09345689084794785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,10240,0.19176444742414686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,4096,0.021822222405009802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,3584,0.01959466603067186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,3072,0.017850667238235474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,3584,0.06678577926423815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,4096,0.07647733555899726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,2560,0.01517688896920946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,65536,1.19869507683648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,2048,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,1536,0.010663111176755695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,3072,0.0583217806286282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,1024,0.00835111074977451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,2048,0.04092622134420607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,1536,0.0324942237801022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,768,0.0075128889746136135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,2560,0.050548444191614784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,512,0.006314666734801398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,1024,0.02513422237502204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,256,0.0052266667286554975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,128,0.004609777695602841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,768,0.02181688944498698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,64,0.004291555533806483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,256,0.015492444237073263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,2048,32,0.004558222161398994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,128,0.014890667464998035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,512,0.01828266680240631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,64,0.014499555031458536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,2048,32,0.01480888823668162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,65536,0.21904799673292372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,51200,0.17636888557010225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,16384,0.06068533658981323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,12288,0.04817777872085571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,16384,0.29154133796691895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,10240,0.039232889811197914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,12288,0.2200017770131429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,8192,0.03085511260562473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,10240,0.18332178062862822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,51200,0.8927058113945855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,8192,0.1426533328162299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,7168,0.028378665447235107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,65536,1.1402044296264648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,6144,0.02420799930890401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,5120,0.02092444399992625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,7168,0.1239395538965861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,4096,0.01752355529202355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,6144,0.10595022307501899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,5120,0.08828800254397923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,3584,0.016199999385409884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,3072,0.01442399952146742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,4096,0.07217599948247273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,2560,0.012727110750145383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,3584,0.06413066387176514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,2048,0.011168888873524137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,2560,0.047001779079437256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,1536,0.0087333329849773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,2048,0.03810666667090522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,3072,0.05516533388031853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,1024,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,768,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,1536,0.031268444326188825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,512,0.005312888986534542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,768,0.021026665965716045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,256,0.004552000098758274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,1024,0.024193776978386775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,512,0.017596445149845548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,128,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,256,0.014834667245546976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,64,0.004607111215591431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,128,0.014130666851997375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1536,32,0.004791111167934206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,64,0.013599110974205865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1536,32,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,65536,0.15804978211720785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,51200,0.12461422549353705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,16384,0.045351998673544995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,16384,0.2721804512871636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,12288,0.032560888263914324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,51200,0.8242408964369031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,65536,1.049282709757487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,10240,0.027709333433045283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,12288,0.2059884468714396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,8192,0.023408000667889912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,7168,0.021017778250906203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,10240,0.17064978016747367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,6144,0.019550222489568923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,8192,0.13369066185421413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,5120,0.016304888659053378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,6144,0.09799200296401978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,4096,0.014426666829321118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,7168,0.11566221714019775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,5120,0.0818915565808614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,3584,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,3072,0.011071110765139261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,4096,0.06723555591371325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,3584,0.059303998947143555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,2560,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,2048,0.009758222434255812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,2560,0.04378044605255127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,3072,0.05156977640257942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,1536,0.007700444095664554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,2048,0.035805334647496544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,1024,0.006005333529578314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,768,0.005408000200986862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,1024,0.022592888938056097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,512,0.004876444323195351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,768,0.020063110523753695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,1536,0.0292195545302497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,256,0.004611555486917496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,128,0.00423911131090588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,512,0.016495111915800307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,64,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,128,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,256,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,1024,32,0.003970666478077571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,64,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,65536,0.1338933308919271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,51200,0.10567822058995564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,1024,32,0.013235555754767524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,16384,0.03706399930848016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,16384,0.2636355559031169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,12288,0.02685244381427765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,51200,0.7933386696709527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,10240,0.022364444202846948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,10240,0.1648319959640503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,12288,0.1994311147265964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,8192,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,7168,0.016576889488432143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,7168,0.11139644516838922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,8192,0.12909777959187826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,6144,0.01481599940194024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,5120,0.013174222575293647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,65536,1.009968015882704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,4096,0.010809777511490716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,5120,0.07892533143361409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,6144,0.09460710816913181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,4096,0.06445066796408759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,3072,0.009020444419648912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,3584,0.056445333692762584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,2560,0.008405333591832055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,2048,0.007265778051482306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,3072,0.04840977655516731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,2048,0.03438399897681342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,2560,0.04086577892303467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,1536,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,1024,0.0063688887490166565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,1024,0.021920889616012573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,1536,0.02826311190923055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,768,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,512,0.00490399988161193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,512,0.01588888963063558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,768,0.019440000255902607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,256,0.004601777841647466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,128,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,128,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,256,0.014147554834683737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,64,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,32,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,32,0.012395555774370829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,768,64,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,768,3584,0.010113777385817634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,65536,0.10024622413847183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,51200,0.08072711361779107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,16384,0.03192799952295091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,16384,0.253219551510281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,12288,0.02514755560292138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,51200,0.7605582343207465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,10240,0.024193776978386775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,65536,0.9723013771904839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,12288,0.19114932749006483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,8192,0.018174222773975797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,7168,0.016889777448442247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,10240,0.15843289428287083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,6144,0.015770667129092746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,8192,0.12427732679578994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,5120,0.0120666664507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,6144,0.0908577773306105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,7168,0.10742311345206366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,4096,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,5120,0.07595022519429524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,3584,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,3072,0.010427555276287926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,3584,0.05393777622116936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,4096,0.0621431138780382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,2560,0.00831288927131229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,2048,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,2048,0.032272001107533775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,2560,0.039176887936062284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,1536,0.005903111149867375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,1024,0.0053013331360287136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,1536,0.027170666389995154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,1024,0.021067554752031963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,768,0.004880889008442561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,512,0.004600888739029567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,768,0.01886399918132358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,256,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,512,0.015600000818570455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,3072,0.04647377795643277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,128,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,256,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,64,0.003588444242874781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,512,32,0.0032764443506797156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,128,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,64,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,65536,0.06728977627224393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,51200,0.06012355619006687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,512,32,0.012496000362767113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,16384,0.021721776988771226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,16384,0.24263289239671496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,65536,0.9292942682902018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,51200,0.729099538591173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,12288,0.01724799970785777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,10240,0.014800889624489678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,8192,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,12288,0.18381688329908583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,7168,0.011377777490350934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,8192,0.11895910898844402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,6144,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,10240,0.1525591082043118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,5120,0.009009777671760982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,7168,0.10210133261150783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,4096,0.007965332931942409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,5120,0.07287910911771986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,3584,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,3072,0.0070773329999711775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,3072,0.04383377896414863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,3584,0.05176266696718004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,2560,0.0063475556671619415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,2048,0.005991111199061076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,2560,0.03741866681310866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,2048,0.03094666533999973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,6144,0.08662844366497463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,1536,0.00499199993080563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,1024,0.00423555572827657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,768,0.0038782221575578055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,1536,0.025782222549120586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,512,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,4096,0.05896711349487305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,1024,0.02079288827048408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,768,0.018210666047202218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,128,0.0031644445326593188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,512,0.01518133282661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,256,0.012768000364303589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,128,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,64,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,32,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,64,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,65536,0.050399998823801674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,256,32,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,51200,0.04166222280926175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,256,256,0.003400000019205941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,16384,0.014135999812020196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,16384,0.24217422803243002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,65536,0.9288995530870227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,12288,0.012833777401182385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,51200,0.7289271354675293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,10240,0.01181688904762268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,8192,0.010794666906197866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,12288,0.18356000052558052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,7168,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,8192,0.11824799908532037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,10240,0.15287555588616267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,6144,0.009721777505344814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,5120,0.008606221940782335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,7168,0.10148000054889256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,4096,0.007636444436179266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,5120,0.07216533025105794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,6144,0.08662133084403144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,3584,0.007000000112586551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,4096,0.058303111129336886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,3072,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,2560,0.00590133335855272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,3072,0.04310399956173367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,3584,0.051206221183141075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,2048,0.0053048887186580235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,2560,0.03639466563860575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,1536,0.004568888909286923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,1024,0.00416711096962293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,1536,0.025747555825445387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,768,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,1024,0.02012622190846337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,2048,0.03092977735731337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,512,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,256,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,768,0.01788444485929277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,128,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,512,0.015198222464985318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,64,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,128,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,128,32,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,64,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,32,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,65536,0.049401776658164136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,51200,0.04003466831313239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,128,256,0.012135110795497894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,16384,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,65536,0.9282550811767578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,12288,0.011962666279739805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,51200,0.7285795741611056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,10240,0.011123555401961008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,16384,0.24214045206705728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,8192,0.009899555808968013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,12288,0.18316533830430773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,7168,0.009313777916961247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,10240,0.15255378352271184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,8192,0.11851910750071208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,6144,0.008736000292830998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,7168,0.1014817754427592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,5120,0.008644444247086843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,4096,0.007608888877762689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,6144,0.08611466487248738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,3584,0.006963555183675554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,5120,0.07182400094138251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,3072,0.006258666515350342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,3584,0.050401777029037476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,4096,0.05903377797868517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,2560,0.00564444437623024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,3072,0.04348088966475593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,2048,0.0052577778697013855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,1536,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,2048,0.030604445272021826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,2560,0.03674044542842441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,1024,0.0041759999261962045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,1536,0.025438222620222304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,768,0.0037644443412621817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,512,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,768,0.01789955629242791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,256,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,512,0.014852444330851236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,1024,0.020447111792034574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,128,0.0028275555620590844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,64,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,256,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,64,32,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,128,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,64,0.011425777441925473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,65536,0.04626489016744825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,51200,0.03810933232307434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,16384,0.010084444450007545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,16384,0.24217867851257324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,65536,0.9282995859781901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,12288,0.008714666797055138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,51200,0.7300702200995551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,10240,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,64,32,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,12288,0.18670222494337294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,8192,0.013212444053755866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,10240,0.15433422724405924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,7168,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,6144,0.010689777632554373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,7168,0.10133510828018188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,8192,0.12232977814144558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,5120,0.010590222146775989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,6144,0.08675199747085571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,4096,0.00925244473748737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,3584,0.00849599970711602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,4096,0.05943466557396782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,3072,0.007622222105662028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,3584,0.051183998584747314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,2560,0.006940444310506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,3072,0.04468533396720886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,2048,0.004986666556861666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,2560,0.03611288799179925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,1536,0.0047048889100551605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,2048,0.030590222941504583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,1024,0.003944000022278892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,1536,0.025736000802781846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,768,0.003551111039188173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,1024,0.020017777880032856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,512,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,5120,0.07430577940411039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,256,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,512,0.014513777361975776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,128,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,256,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,128,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,64,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2049,32,32,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,64,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,32,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2049,32,768,0.017966222431924608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,51200,6.435098859998916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,16384,2.1024142371283636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,16384,2.866594738430447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,51200,9.882982889811197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,16384,2.138640933566623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,12288,1.559785736931695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,12288,2.027803632948134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,10240,1.1975057390001085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,51200,6.745200686984592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,8192,0.986439069112142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,12288,1.6144222683376734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,10240,1.6529137293497722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,10240,1.3334791395399306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,8192,1.294416003757053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,8192,1.0790977478027344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,7168,1.1178515752156575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,7168,0.8245431052313911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,6144,0.7870231204562717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,7168,0.9446008470323352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,6144,0.9453555213080512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,5120,0.5831439759996202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,5120,0.7882711092631022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,6144,0.8150800069173177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,4096,0.5213662253485786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,4096,0.6273591253492568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,3584,0.4418906635708279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,5120,0.6916249063279895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,3072,0.4036240047878689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,4096,0.5673111279805502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,3584,0.5504755443996853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,2560,0.29799021614922416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,3584,0.5013093418545193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,3072,0.5233919885423448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,2048,0.2525431050194634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,2560,0.4077742099761963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,2560,0.3766888777414958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,2048,0.33431466420491535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,1536,0.19700445069207084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,2048,0.3169164392683241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,1536,0.26487112045288086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,1536,0.2546488973829481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,1024,0.20031910472446016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,3072,0.4434062110053168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,1024,0.19471555285983613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,768,0.10372710890240139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,512,0.08054400152630277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,768,0.16375733746422663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,512,0.13687645064459905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,256,0.060772445466783315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,768,0.16199289427863225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,512,0.13343377908070883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,128,0.05223911007245382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,1024,0.14547822210523817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,256,0.12208622031741673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,64,0.04531822270817227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,256,0.10734399822023179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,128,0.11653955777486165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,65536,128,0.08513955275217693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,65536,32,0.04815466536415947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,64,0.11734933323330349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,65536,32,0.1180444426006741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,65536,6.108826531304254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,65536,9.890496148003473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,51200,4.917676289876302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,65536,7.005455864800347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,16384,1.6136648390028212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,16384,2.307048797607422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,16384,1.7480764389038086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,12288,1.2240924835205078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,51200,7.881293402777778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,10240,0.9845102098253039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,12288,1.6031102074517145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,12288,1.306985749138726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,10240,1.319230185614692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,51200,5.512810601128472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,8192,0.854823112487793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,10240,1.1010391447279189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,8192,1.032298617892795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,7168,0.7085084385342069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,7168,0.8892435497707791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,6144,0.5906169149610732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,8192,0.8800906605190701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,7168,0.7781964408026801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,6144,0.7569537692599826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,5120,0.5296426879035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,6144,0.6709982024298774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,5120,0.6308719846937392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,4096,0.4574426545037164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,5120,0.5683884620666504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,3584,0.3580008877648248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,4096,0.504841751522488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,3072,0.31396712197197807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,3584,0.4416791068183051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,4096,0.46623823377821183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,3584,0.4156186580657959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,2560,0.251656002468533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,3072,0.3899502224392361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,3072,0.3647155496809218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,2048,0.2065368890762329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,2560,0.32852888107299805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,2560,0.3117662270863851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,2048,0.2692320081922743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,1536,0.16078844335344103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,2048,0.26182399855719674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,1024,0.11745244926876491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,1536,0.21361777517530653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,768,0.09030133485794067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,1536,0.21112799644470215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,1024,0.16178045007917616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,1024,0.1622426642311944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,512,0.06422666708628337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,768,0.13042844666375056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,768,0.13664178053538004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,256,0.06166488594479031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,512,0.1119004487991333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,128,0.04186488853560554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,256,0.09866311152776082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,256,0.08961777554617988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,64,0.04010044535001119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,512,0.11088711023330688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,51200,128,0.07115644216537476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,128,0.09391822417577107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,51200,32,0.04226488868395487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,64,0.09494400024414062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,51200,32,0.09493066867192586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,51200,1.7507919735378688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,65536,2.206287172105577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,65536,3.6984452141655813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,16384,0.5677013397216797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,65536,3.0977475908067493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,51200,2.9499751196967234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,16384,0.8507111337449816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,12288,0.432370662689209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,51200,2.4263803693983292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,12288,0.6159697638617622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,16384,0.79366668065389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,10240,0.3136471112569173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,12288,0.6018053160773383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,8192,0.26259999805026585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,10240,0.507913801405165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,8192,0.3981226550208197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,10240,0.506330649058024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,7168,0.222053329149882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,8192,0.402366214328342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,6144,0.19994844330681694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,7168,0.34787999259101016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,5120,0.16157066822052002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,7168,0.3555973370869954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,6144,0.29801689253913033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,6144,0.30659911367628306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,5120,0.24872178501553008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,5120,0.2600897683037652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,3584,0.11933422088623047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,4096,0.2009119987487793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,4096,0.2142497830920749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,3072,0.10572977860768636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,3584,0.17726311418745253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,3584,0.19188711378309461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,3072,0.1550648874706692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,2560,0.09098044368955825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,3072,0.16962667306264242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,4096,0.1359928846359253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,2560,0.13389600647820368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,2048,0.067449775007036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,2560,0.14590666029188368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,1536,0.05351555678579542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,2048,0.11073688666025798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,2048,0.12433244122399224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,1536,0.08887377712461685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,1024,0.03794577717781067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,1536,0.10235466559727986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,768,0.030822220775816176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,1024,0.06780888636906941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,1024,0.07928711175918579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,512,0.021983999345037673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,768,0.055461333857642285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,768,0.06885511345333524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,256,0.01829777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,512,0.05750755469004313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,512,0.04615555538071526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,128,0.017501332693629794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,256,0.040192888842688665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,256,0.04685777756902906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,64,0.018583110637134977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,128,0.03738488753636678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,16384,32,0.01919822229279412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,16384,128,0.03940977652867635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,64,0.03650310966703627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,16384,32,0.035624000761244036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,65536,1.638228416442871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,51200,1.271172417534722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,65536,3.048166275024414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,65536,2.705069435967339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,16384,0.4374355475107829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,16384,0.6863102383083768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,51200,2.3683431413438583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,12288,0.32468001047770184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,51200,2.1272320217556424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,16384,0.6941466861300998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,10240,0.25897155867682564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,12288,0.5271573596530491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,12288,0.4999137984381781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,8192,0.2184542285071479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,10240,0.41431911786397296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,7168,0.1911182271109687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,10240,0.4423084523942735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,8192,0.3529226779937744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,8192,0.3272284401787652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,6144,0.15986400180392796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,7168,0.2846711211734348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,5120,0.13624355528089735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,7168,0.31038043234083385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,6144,0.24403021070692274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,6144,0.26788534058464897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,5120,0.2041591140958998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,4096,0.10156355301539104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,3584,0.08916711144977146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,4096,0.16488888528611925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,4096,0.18680177794562447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,3072,0.08035822047127618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,3584,0.16751555601755777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,3584,0.14680088890923396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,2560,0.06519999768998888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,3072,0.12867289119296604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,3072,0.1487119992574056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,2048,0.05568799707624647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,2560,0.12785777780744764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,2560,0.11130932966868083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,1536,0.04124355647299025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,2048,0.09170755412843491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,2048,0.10925066471099854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,1024,0.03035822179582384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,1536,0.07344888978534274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,1536,0.09041333198547363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,768,0.023964444796244305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,1024,0.055508444706598915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,1024,0.07065866390864055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,768,0.04501510990990532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,512,0.019211555520693462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,768,0.06160000297758314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,256,0.014477334088749357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,512,0.036738667223188616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,512,0.052248888545566134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,256,0.031707555055618286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,128,0.01368711143732071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,128,0.029806223180558946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,64,0.011360000405046674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,64,0.02956177790959676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,128,0.03479822145568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,12288,32,0.030044443077511255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,256,0.04009244508213467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,12288,5120,0.22697154680887857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,12288,32,0.011713777979214987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,65536,1.3344258202446833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,51200,1.0207288530137804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,65536,2.664093229505751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,16384,0.32665689786275226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,51200,2.0643030802408853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,65536,2.452998267279731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,16384,0.620006243387858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,12288,0.25375021828545463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,16384,0.6310293409559462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,51200,1.9189688364664714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,12288,0.4574417538113064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,10240,0.2395519945356581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,12288,0.47910224066840273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,8192,0.1894995503955417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,10240,0.3691706657409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,10240,0.40201687812805176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,7168,0.16672355598873564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,8192,0.29083021481831867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,8192,0.3199102348751492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,7168,0.25279466311136883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,6144,0.14090222782558864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,7168,0.28143466843499076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,6144,0.21608000331454805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,6144,0.24286222457885742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,5120,0.11656088299221462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,4096,0.09416621923446655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,5120,0.18066488371955025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,5120,0.20621689160664877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,3584,0.08630844619539048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,4096,0.14775466918945312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,4096,0.16987644301520455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,3072,0.07162044445673625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,3584,0.13063200314839682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,3584,0.15334578355153403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,3072,0.13508799340989855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,2560,0.06064177883995903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,3072,0.11491466893090142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,2048,0.04879822333653768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,2560,0.09957244661119248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,2560,0.11784799893697102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,2048,0.08281422323650785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,1536,0.03917510973082648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,2048,0.10018577840593125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,1024,0.029176000091764662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,1536,0.06605777475568983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,1536,0.08340444167455037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,768,0.022115555074479844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,1024,0.06474488973617554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,768,0.05715644359588623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,512,0.01739911072784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,768,0.04063644342952304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,512,0.033435556623670794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,256,0.013263111313184103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,512,0.048487110270394214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,256,0.027785778045654297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,128,0.013078221844302284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,256,0.039729777309629656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,1024,0.04866933491494921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,128,0.026854222019513447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,64,0.011388444238238864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,10240,32,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,64,0.026802667313151892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,10240,32,0.026398221651713055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,10240,128,0.03228177626927694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,65536,1.1147901746961806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,51200,0.8488364219665527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,65536,2.2998133765326605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,16384,0.2908044391208225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,16384,0.534173329671224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,51200,1.7741333643595378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,65536,2.1895813412136502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,12288,0.21789956092834473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,16384,0.5659644338819716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,10240,0.18428444862365723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,51200,1.7180275387234156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,12288,0.3902382320827908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,12288,0.4300311141543918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,8192,0.14976355764600965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,10240,0.3234577708774143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,7168,0.13234578238593206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,10240,0.36189688576592344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,8192,0.2542791101667616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,6144,0.11040888892279731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,8192,0.2878257698482937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,7168,0.2204915550020006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,7168,0.252454227871365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,5120,0.09290133582221137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,6144,0.18905956215328643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,4096,0.07574933105044894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,6144,0.21805066532558867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,5120,0.15749866432613796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,5120,0.18468177318572998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,3584,0.06722399923536512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,4096,0.12815733750661215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,4096,0.15303556124369302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,3072,0.05690755446751913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,3584,0.11501510938008626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,3584,0.13726045025719538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,2560,0.04840266704559326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,3072,0.12220888667636448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,3072,0.10168711344401042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,2560,0.08731200297673543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,2048,0.03587199913130866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,2560,0.10610222154193455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,1536,0.026679999298519556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,2048,0.09085955884721543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,1536,0.05702844593260023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,1536,0.07527822256088257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,2048,0.07315022415584989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,1024,0.043142222695880465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,1024,0.059559113449520536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,768,0.017844445175594754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,768,0.052339557144376964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,512,0.015215999550289579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,768,0.03677689035733541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,512,0.029245333539115057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,256,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,512,0.044235554006364614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,1024,0.02117777698569828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,256,0.024700444605615404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,128,0.01034755590889189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,128,0.023778667052586872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,128,0.02934755550490485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,8192,256,0.03808000021510654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,32,0.010855110983053843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,64,0.023415999280081853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,8192,32,0.023775110642115276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,65536,0.8748639954460992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,8192,64,0.010286221901575724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,51200,0.6844159762064616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,65536,2.092150158352322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,51200,1.6439796023898656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,16384,0.2248684565226237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,65536,2.0634471045600042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,16384,0.48832088046603733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,12288,0.16616800096299914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,16384,0.5341591305202907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,51200,1.617690616183811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,10240,0.15768355793423122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,12288,0.36205066574944395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,12288,0.40686578220791286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,8192,0.11919732888539632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,10240,0.34061243798997665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,10240,0.30107466379801434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,7168,0.10592177841398452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,8192,0.23724444707234701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,8192,0.27064355214436847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,6144,0.0890257755915324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,7168,0.23817510075039336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,7168,0.2061742279264662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,5120,0.07513955566618177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,6144,0.17596088515387642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,4096,0.06033600038952298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,5120,0.1473697821299235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,5120,0.17475910981496176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,4096,0.11955644024742974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,3584,0.054125332170062594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,4096,0.14423378308614096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,3072,0.047784000635147095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,3584,0.10648444626066421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,3584,0.1304524474673801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,3072,0.09447555409537421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,3072,0.11552711327870686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,2560,0.039507554637061224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,2048,0.03129511078198751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,2560,0.08169866932762994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,6144,0.20556978384653726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,1536,0.025755556093321905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,2048,0.08588533269034491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,2048,0.06866311364703707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,1024,0.01811022228664822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,1536,0.05360888772540622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,1536,0.07145244545406766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,768,0.014890667464998035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,1024,0.05640977621078491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,1024,0.04037244452370538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,768,0.0333280000421736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,512,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,768,0.05017422305213081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,256,0.009009777671760982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,512,0.026055999928050574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,512,0.04224710994296604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,128,0.008368888662921058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,256,0.023123555713229712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,2560,0.10021155410342747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,256,0.03416000141037835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,64,0.007274666594134436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,7168,32,0.0075911109646161394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,7168,128,0.028861333926518757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,64,0.021996445126003687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,128,0.022666666242811415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,7168,32,0.022030222747060988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,51200,0.6492231157090929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,65536,0.8445670869615344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,65536,1.9347750345865886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,16384,0.21838933891720244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,51200,1.498429298400879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,65536,2.0600836012098527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,16384,0.45089244842529297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,51200,1.6168728934393988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,12288,0.16853866312238905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,16384,0.53238402472602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,10240,0.14261066913604736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,12288,0.3347511026594374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,8192,0.10972088575363159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,10240,0.27831734551323783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,12288,0.4052942328982883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,10240,0.34024887614780003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,8192,0.21821510791778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,7168,0.09859999683168198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,8192,0.2698044512006971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,6144,0.08943733241822983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,7168,0.18939288457234701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,7168,0.23692711194356283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,6144,0.16216888692643908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,6144,0.20372977521684435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,5120,0.0708728896247016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,4096,0.05689422289530436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,5120,0.135935107866923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,5120,0.17347199387020537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,3584,0.05215555429458618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,4096,0.11159822675916885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,4096,0.14337066809336343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,3072,0.04116977916823493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,3584,0.09931821955574884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,3584,0.12948266665140787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,3072,0.11434222592247857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,2560,0.03621777892112732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,2560,0.07603110869725545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,2048,0.028184887435701158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,2560,0.09943111075295342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,2048,0.06183911032146878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,1536,0.02258666687541538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,2048,0.08523555596669515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,1536,0.04909155435032315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,1024,0.016512889001104567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,3072,0.08668444554011027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,1536,0.07103466987609863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,768,0.014205333259370593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,1024,0.037111110157436795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,1024,0.05548977851867676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,512,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,768,0.04869777626461453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,768,0.03235910998450385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,512,0.025228444072935317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,256,0.01034311122364468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,512,0.03965955641534593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,128,0.009906666974226633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,256,0.02168888847033183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,256,0.03331999977429708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,64,0.0094373333785269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,6144,128,0.028363555669784546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,6144,32,0.00959911114639706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,64,0.02072000006834666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,32,0.02035999960369534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,65536,0.7750275399949816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,6144,128,0.020623111062579684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,51200,0.5812062157524956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,65536,1.7317484749688044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,51200,1.375825775994195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,16384,0.19274044036865234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,65536,1.9297883775499132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,16384,0.41354576746622723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,51200,1.5136541790432398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,12288,0.145087997118632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,10240,0.11474311351776123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,16384,0.5030390951368544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,12288,0.30843377113342285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,8192,0.08772533469729954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,12288,0.3809679879082574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,10240,0.25655465655856663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,7168,0.07783910963270399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,8192,0.2009751134448581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,10240,0.3205182287428114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,8192,0.2531386746300591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,6144,0.06889333327611287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,7168,0.1747573349210951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,7168,0.2218142218059964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,6144,0.14941155910491943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,6144,0.19201244248284233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,5120,0.12495022349887425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,4096,0.04660355713632372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,4096,0.1020888884862264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,3584,0.040335112147861056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,4096,0.1354124413596259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,3584,0.09081777599122788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,3072,0.03426133261786567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,3584,0.12203733126322429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,5120,0.055629332860310875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,2560,0.030442665020624798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,3072,0.07867822382185194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,2560,0.06576533450020684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,2560,0.09336088763342963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,5120,0.16301688883039686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,2048,0.02475111186504364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,3072,0.10853333605660333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,1536,0.020190222395790946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,2048,0.08009866873423259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,2048,0.0566248893737793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,1024,0.0148053334818946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,1536,0.04547199938032362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,1024,0.03417511118782891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,1024,0.05161066518889534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,768,0.028620445066028174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,768,0.045647998650868736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,512,0.00907022257645925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,512,0.02269688910908169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,512,0.03800444470511542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,256,0.007324444750944774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,256,0.020296888218985665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,1536,0.06717866659164429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,256,0.03125866585307651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,128,0.0069671107663048645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,768,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,5120,128,0.027591110931502447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,128,0.019262222780121695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,64,0.006355555521117316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,5120,32,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,64,0.018992889258596633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,5120,32,0.019003555178642273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,65536,0.5477866596645778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,51200,0.4290640089246962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,65536,1.5594453811645508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,51200,1.217375119527181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,16384,0.14559733867645264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,65536,1.8014310201009114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,16384,0.3773377736409505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,12288,0.10910133520762126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,16384,0.4692844284905328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,51200,1.4134773678249781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,12288,0.2821528911590576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,10240,0.09761155313915676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,12288,0.35746489630805117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,8192,0.07730044258965386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,10240,0.23411732249789766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,10240,0.2991502285003662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,7168,0.06723822487725152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,8192,0.23704621526930067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,8192,0.18420533339182535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,7168,0.16020445028940836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,6144,0.05802222092946371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,7168,0.2069573269950019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,5120,0.048432889911863536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,6144,0.1362560060289171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,6144,0.17974400520324707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,4096,0.04038577940728929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,5120,0.11510666211446126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,5120,0.1532764434814453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,3584,0.03605688942803277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,4096,0.12677510579427084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,4096,0.09411111142900254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,3584,0.08342578013737996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,3072,0.031573332018322416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,3584,0.11378310786353217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,2560,0.026536888546413843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,3072,0.07190755340788099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,3072,0.10162488619486491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,2560,0.06084711021847195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,2048,0.02272622287273407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,2048,0.0503475566705068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,1536,0.018035555879275005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,2048,0.07561510801315308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,1536,0.04103022151523166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,1536,0.06372000111473931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,1024,0.0313368903266059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,768,0.011160888605647616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,2560,0.08781422509087457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,1024,0.0484168893761105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,512,0.00869511150651508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,768,0.02611644400490655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,768,0.04356177647908529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,256,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,512,0.03568711214595371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,512,0.021754667162895203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,256,0.018262222409248352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,128,0.006647999915811751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,1024,0.01347733371787601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,256,0.030020445585250854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,64,0.006959111326270633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,4096,32,0.007282666862010956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,128,0.017737777696715463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,64,0.01755288905567593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,4096,128,0.025570667452282373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,4096,32,0.01755644381046295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,65536,0.5266453425089518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,51200,0.4077377849155002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,65536,1.483112865024143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,16384,0.13930400212605795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,51200,1.1557191212972004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,16384,0.3586720095740424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,51200,1.313719113667806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,12288,0.10441599951850043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,12288,0.26940711339314777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,10240,0.0943031112353007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,12288,0.3329582214355469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,10240,0.2225511074066162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,10240,0.2791324456532796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,8192,0.07631288634406196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,65536,1.6748586230807836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,8192,0.17487111356523302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,16384,0.4371831152174208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,7168,0.06663378079732259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,8192,0.22076266341739229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,6144,0.058840003278520375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,7168,0.15151466263665095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,7168,0.19413688447740343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,6144,0.12974133756425646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,5120,0.048892445034450956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,6144,0.16771911250220406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,5120,0.10884622070524429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,4096,0.04023199942376878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,5120,0.14262133174472386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,3584,0.035820444424947105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,4096,0.08885777658886379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,4096,0.11879911687639023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,3072,0.03134311238924662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,3584,0.10662844445970322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,3072,0.0949359999762641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,2560,0.027425777581002977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,3072,0.07067288955052693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,2560,0.05917155742645264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,2048,0.02232088810867733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,2560,0.08290844493442111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,2048,0.04766044351789686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,1536,0.01796266602145301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,2048,0.07154044177797106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,1536,0.03910133242607117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,1536,0.05985333522160848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,1024,0.02924622098604838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,1024,0.045511109961403735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,768,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,3584,0.07865955432256062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,768,0.04116355710559421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,768,0.02546311087078518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,512,0.008707555631796518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,512,0.03525422347916497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,256,0.006784889019197888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,512,0.020637333393096924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,256,0.01774222155412038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,256,0.029571556382709082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,128,0.016943999462657504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,1024,0.013532444834709167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,64,0.0052888890107472735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3584,128,0.025506666964954797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,32,0.00573333352804184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,64,0.016943999462657504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3584,32,0.016919111212094624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3584,128,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,65536,0.4233040014902751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,51200,0.32233066029018825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,65536,1.3896915647718642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,51200,1.0903298060099285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,16384,0.111135999361674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,65536,1.6711040072970922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,16384,0.3397493362426758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,12288,0.08439555433061387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,16384,0.4360026783413357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,12288,0.25625867313808864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,10240,0.06776355372534858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,51200,1.3138977686564128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,12288,0.3322693241967095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,8192,0.057040890057881675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,10240,0.21314666006300184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,7168,0.046909332275390625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,10240,0.2796497874789768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,8192,0.21984622213575575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,8192,0.16765688525305855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,6144,0.04359733396106296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,7168,0.14523555172814265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,5120,0.03464711043569777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,6144,0.12414844830830891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,7168,0.19332533412509492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,6144,0.16668089230855307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,4096,0.03029777937465244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,5120,0.1043155524465773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,3584,0.025615110993385315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,5120,0.14221599366929796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,4096,0.08598755465613471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,3072,0.022703111171722412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,3584,0.0766577786869473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,4096,0.1186479992336697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,3584,0.10600977473788792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,2560,0.020437333318922255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,3072,0.06704888741175334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,2048,0.016755556066830952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,2560,0.05664622121387058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,2560,0.08278844753901164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,1536,0.014027555783589682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,2048,0.04534133275349935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,2048,0.07029955254660712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,1024,0.010265777508417765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,1536,0.05925333499908447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,1536,0.03718577822049459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,1024,0.02843911117977566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,768,0.008636444807052612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,1024,0.04488266507784525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,512,0.007290666302045186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,768,0.024121777878867254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,512,0.019592000378502738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,256,0.006761777732107375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,512,0.03400977783732944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,3072,0.09462400277455647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,256,0.01683466633160909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,256,0.029354665014478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,128,0.025566221939192876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,64,0.005919111271699269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,64,0.0162435554795795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,3072,768,0.04018933243221707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,32,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,3072,128,0.006573333508438534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,32,0.016220443778567843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,3072,128,0.016135111451148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,65536,0.3671528763241238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,51200,0.285748455259535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,65536,1.2978533638848198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,51200,1.0144009060329862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,16384,0.09795644548204209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,65536,1.671681827969021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,16384,0.3226168950398763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,12288,0.07346844673156738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,16384,0.4358551237318251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,12288,0.24362754821777344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,51200,1.3134364022148979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,10240,0.056794669893052846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,12288,0.3321502208709717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,8192,0.04787733488612705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,10240,0.27951733271280926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,10240,0.2023226684994168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,7168,0.04038222299681769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,8192,0.15902666250864664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,8192,0.22102399667104086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,6144,0.03600000010596381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,7168,0.1925039953655667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,7168,0.13847111331091985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,5120,0.029631111356947158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,6144,0.1174631118774414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,6144,0.16636178228590223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,4096,0.025214221742418077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,5120,0.14155111047956678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,5120,0.09851466947131687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,3584,0.022566222482257422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,4096,0.08149777518378364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,4096,0.11781688531239827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,3072,0.01978488928741879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,3584,0.10550577772988214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,3584,0.07258933120303683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,2560,0.016562667157914903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,3072,0.06285333633422852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,3072,0.09401333332061768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,2048,0.01460444430510203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,2560,0.08190400070614286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,2560,0.0529386666085985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,2048,0.04344355397754245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,1536,0.011671110987663269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,2048,0.06960711214277479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,1024,0.009093333449628618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,1536,0.03445600138770209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,1536,0.05845599704318576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,1024,0.02664622167746226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,768,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,1024,0.04415200153986613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,512,0.006826666494210561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,768,0.02268444498380025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,768,0.04014755619896783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,512,0.01881866653760274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,256,0.005949333310127258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,512,0.033655110332700945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,128,0.005274666680230035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,256,0.016156444946924847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,256,0.02823644545343187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,64,0.005314666777849197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,128,0.015461333923869662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2560,128,0.02375288969940609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2560,32,0.0052782222628593445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,32,0.0148026661740409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2560,64,0.015488000379668342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,65536,0.28848444090949166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,51200,0.22353866365220812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,65536,1.2071146435207791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,16384,0.07959466510348849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,51200,0.9461608462863498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,16384,0.3046017752753364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,65536,1.5447360144721136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,51200,1.211045371161567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,12288,0.056779556804233126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,16384,0.40435110198126895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,10240,0.048997332652409874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,10240,0.1905857721964518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,12288,0.30788445472717285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,10240,0.2584151162041558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,8192,0.039431111680136785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,7168,0.03545066714286804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,8192,0.14994666311475965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,8192,0.20370311207241484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,7168,0.12952178054385716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,6144,0.03138044476509094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,7168,0.1784151130252414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,5120,0.026731555660565693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,6144,0.11001422007878621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,6144,0.1537733342912462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,4096,0.022388445006476507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,5120,0.09258577558729385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,5120,0.13171554936303034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,3584,0.02033866610791948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,4096,0.07630222373538546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,12288,0.22984088791741264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,4096,0.10896355575985378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,3072,0.018034666776657104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,3584,0.06741155518425836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,3072,0.08797155486212836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,3072,0.05851733022265964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,2560,0.049970666567484535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,2560,0.0767600006527371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,2048,0.013491555220550962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,3584,0.09859110911687215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,2048,0.0652631123860677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,2048,0.04077422287729051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,2560,0.016114667057991028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,1536,0.010886222124099731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,1024,0.008743999732865227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,1536,0.055594669448004834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,1536,0.033322665426466204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,1024,0.025098666548728943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,768,0.0073831114504072405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,1024,0.04179822074042427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,768,0.02163822286658817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,512,0.006324444380071428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,768,0.03807910945680406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,256,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,512,0.03210577699873183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,512,0.017909333109855652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,128,0.004962666581074397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,256,0.027232888672086928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,128,0.014807111687130399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,256,0.015936000479592215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,64,0.004847110973464118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,2048,32,0.004553777890072929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,2048,128,0.022960889670583937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,32,0.014479999740918478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,65536,0.2351306544409858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,65536,1.1331555048624675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,2048,64,0.014855111638704935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,51200,0.1825582186381022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,65536,1.5437688827514648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,16384,0.0631857779290941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,51200,0.8910338083902994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,16384,0.29066043429904515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,12288,0.04821244544453091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,51200,1.210462252298991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,12288,0.21958043840196398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,16384,0.40497689776950413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,10240,0.04096800088882446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,12288,0.3075973457760281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,10240,0.2593822214338514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,8192,0.033737778663635254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,10240,0.1825573311911689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,7168,0.02953688965903388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,8192,0.14237244923909506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,8192,0.20357333289252388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,6144,0.02564977771706051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,7168,0.12379911210801865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,7168,0.17905599541134307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,6144,0.10510044627719456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,5120,0.02259911100069682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,6144,0.15398489104376897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,5120,0.08799999952316284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,5120,0.13130666149987116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,4096,0.018619555566045973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,3584,0.01664355562792884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,4096,0.07200977537367079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,3584,0.06409244404898749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,3072,0.014787554740905762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,3584,0.09822666645050049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,2560,0.012313777373896705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,3072,0.08728444576263428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,3072,0.055960887008243136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,2560,0.04648355642954508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,2048,0.010451555252075195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,2560,0.07609333594640096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,1536,0.008720888859695857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,2048,0.03786933422088623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,2048,0.06545777453316583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,1024,0.007517333659860823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,1536,0.03120000163714091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,1536,0.053210665782292686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,768,0.007937777373525832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,4096,0.10941866371366714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,1024,0.04140799906518724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,512,0.005799110978841782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,1024,0.02405511173937056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,768,0.02090399960676829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,768,0.0374017788304223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,256,0.005031110925806893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,512,0.01687022215790219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,128,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,512,0.031940443648232356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,256,0.014538667268223233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,256,0.02753777801990509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,64,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1536,32,0.0042035554846127825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,128,0.013793778088357715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,64,0.013882666826248169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1536,128,0.024159999357329473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1536,32,0.013786666923099093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,65536,0.16043110688527426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,51200,0.1251022285885281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,51200,0.8217244678073459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,65536,1.0476773579915364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,16384,0.04507288999027676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,51200,1.1101030773586698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,65536,1.4142622417873805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,16384,0.27193689346313477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,12288,0.0329733325375451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,16384,0.3713946607377794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,12288,0.20501422882080078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,10240,0.0266942216290368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,12288,0.28408887651231557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,8192,0.022030222747060988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,10240,0.2387475437588162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,8192,0.18748977449205187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,8192,0.13373422622680664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,7168,0.020768889122539096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,7168,0.16406133439805773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,7168,0.11564089192284478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,6144,0.018048000004556444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,6144,0.14125333891974554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,10240,0.17073867056104872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,5120,0.015602666470739575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,5120,0.08194933334986369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,5120,0.12077421612209743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,4096,0.012994666894276937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,4096,0.06675910949707031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,4096,0.09997333420647515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,3584,0.011809777882364062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,3584,0.05925599733988444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,3584,0.0904186699125502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,3072,0.010371555884679159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,6144,0.09877955251269871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,3072,0.05068177647060818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,3072,0.08058755265341865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,2560,0.009580444130632613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,2048,0.00815466625822915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,2560,0.07027822070651583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,2560,0.04237066706021627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,1536,0.0069982219073507525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,2048,0.03529422150717841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,2048,0.05977866384718153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,1024,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,1536,0.02854222059249878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,1024,0.0226275556617313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,1024,0.03871999846564399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,768,0.0052488889131281115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,768,0.019348444210158456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,768,0.035290665096706815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,512,0.004605333424276776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,512,0.0299271113342709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,256,0.00425511101881663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,256,0.01349511080318027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,256,0.025428444147109985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,128,0.003584889074166616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,1536,0.049478222926457725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,128,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,64,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,1024,128,0.02273955610063341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,1024,32,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,64,0.012874667015340594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,512,0.016155555844306946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,65536,0.1469013293584188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,1024,32,0.013216888739003075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,51200,0.11437866422865127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,65536,1.0096719529893663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,51200,0.7939662403530545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,16384,0.039177778694364764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,65536,1.412753740946452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,12288,0.03020711077584161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,16384,0.2629786597357856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,12288,0.19865689012739393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,16384,0.37091377046373153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,10240,0.026145777768558923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,51200,1.1083457734849718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,12288,0.2830764452616374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,8192,0.02195022172398037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,10240,0.16451377338833278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,10240,0.23809867435031465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,7168,0.019632889164818656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,7168,0.11132089296976726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,8192,0.1865004433525933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,8192,0.12904977798461914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,6144,0.017290666699409485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,7168,0.16338666280110678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,5120,0.01518133282661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,6144,0.09416266944673325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,4096,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,5120,0.07883199718263414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,6144,0.14104089472028944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,5120,0.12036000357733832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,3584,0.011692444483439127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,4096,0.06461422310935126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,4096,0.09958310921986897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,3584,0.056668447123633496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,3072,0.009729777773221334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,2560,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,3584,0.09042844507429336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,3072,0.04818933208783468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,3072,0.08026578028996785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,2048,0.00795022232664956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,2560,0.041434665520985924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,2560,0.0683751106262207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,1536,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,2048,0.033957332372665405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,2048,0.05806399716271294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,1024,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,1536,0.04898577928543091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,1536,0.027886221806208294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,768,0.0053119998839166425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,1024,0.021653332644038733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,1024,0.0381022228135003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,512,0.004950222041871813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,768,0.03457333313094245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,768,0.019133332702848647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,256,0.004611555486917496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,512,0.01587288909488254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,512,0.029862221744325426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,128,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,256,0.025083555115593806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,256,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,64,0.003942222230964237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,128,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,768,128,0.021345777644051447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,768,32,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,65536,0.09098222520616318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,64,0.012849777936935425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,768,32,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,51200,0.07358844412697686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,65536,0.9682942496405708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,51200,0.7607199880811902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,16384,0.02986577815479702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,65536,1.4016141891479492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,16384,0.25253155496385365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,12288,0.02247200078434414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,12288,0.19139733579423693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,16384,0.3686906761593289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,51200,1.0993466907077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,10240,0.01934400035275353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,12288,0.2806302176581489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,8192,0.016545777519543965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,10240,0.15839378039042154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,10240,0.23631376690334746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,7168,0.014723555909262763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,8192,0.12415911091698541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,8192,0.1848400036493937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,6144,0.012591111163298288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,7168,0.10729777812957764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,6144,0.09064355823728774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,5120,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,6144,0.14033155971103245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,5120,0.0759813321961297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,4096,0.009358222285906473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,5120,0.11944622463650173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,4096,0.06192800071504381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,3584,0.00872622181971868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,4096,0.0996204415957133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,3584,0.05447466505898369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,3072,0.007992000215583378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,3584,0.08939466873804729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,3072,0.0458826654487186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,2560,0.007262222468852997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,3072,0.07969511217541166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,2560,0.03902133305867513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,2048,0.006981333096822103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,2560,0.06751466459698148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,2048,0.03222222129503886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,1536,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,2048,0.05968177980846829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,7168,0.1624551084306505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,1536,0.02642755541536543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,1024,0.004953777624501122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,1536,0.04940088921123081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,768,0.004537777768241035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,1024,0.020960888928837244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,1024,0.03804444604449802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,512,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,768,0.03465777635574341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,768,0.01863555610179901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,512,0.015501333607567681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,256,0.003612444632583194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,512,0.029727111260096233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,128,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,256,0.013088000317414602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,256,0.02583733366595374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,64,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,128,0.012339555554919772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,512,128,0.021388444635603163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,512,32,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,65536,0.061426666047838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,64,0.012513777448071374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,512,32,0.01200177768866221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,51200,0.04892444279458788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,51200,0.7290648884243436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,16384,0.0181031111213896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,65536,0.9290249082777234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,16384,0.2422755559285482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,51200,1.1006968816121419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,12288,0.01742844449149238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,65536,1.3996800316704645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,16384,0.366699563132392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,10240,0.014814222852389017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,12288,0.27988889482286244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,10240,0.1526871124903361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,12288,0.18407021628485787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,8192,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,10240,0.23580090204874674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,8192,0.11920800473954941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,7168,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,7168,0.16214222378200954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,6144,0.010066666536860997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,7168,0.10300266742706299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,6144,0.0867733359336853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,5120,0.008933333059151968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,6144,0.1399386723836263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,5120,0.07289244731267293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,4096,0.007649777664078607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,5120,0.11898222234514023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,8192,0.18490133020612928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,3584,0.007302222152551015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,4096,0.05910933017730713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,4096,0.09882755411995782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,3584,0.05099733339415657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,3072,0.006729777902364731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,3584,0.08834399779637654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,2560,0.0063475556671619415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,3072,0.04420800010363261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,3072,0.07944622304704455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,2560,0.0373964442147149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,2048,0.005285333428117964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,2560,0.06825421916113959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,1536,0.004972444640265571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,2048,0.03087644444571601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,2048,0.05844710932837593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,1024,0.004284444368547863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,1536,0.025770665870772466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,1536,0.04801333281728956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,768,0.003896000070704354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,1024,0.0377200014061398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,1024,0.020392888122134738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,768,0.01789955629242791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,512,0.0036346668170558084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,768,0.03432977861828274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,256,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,512,0.014846222268210517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,512,0.029184887806574505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,128,0.0032604444358083936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,256,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,256,0.024547555380397375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,64,0.003156444471743372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,256,128,0.021672889590263367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,128,0.011926222178671094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,256,32,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,64,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,65536,0.05146844519509209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,256,32,0.011872000164455838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,51200,0.04101777738995022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,65536,0.9290071063571506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,51200,0.7284506691826714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,16384,0.012535110943847232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,65536,1.3992302152845595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,16384,0.24230755700005424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,12288,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,16384,0.3669484456380208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,51200,1.0997422536214192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,10240,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,12288,0.1838302214940389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,12288,0.2797386646270752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,8192,0.012059555285506778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,10240,0.1522595617506239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,8192,0.11884621779123943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,7168,0.01075644459989336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,8192,0.1846853362189399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,7168,0.10176533460617065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,7168,0.16217599974738225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,6144,0.08588266372680664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,6144,0.13934755325317383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,5120,0.008949333594905006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,10240,0.23536621199713814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,4096,0.007672000262472365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,5120,0.07208800315856934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,5120,0.11829688813951279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,4096,0.05882933404710558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,4096,0.09791466924879287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,3584,0.007263999846246507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,3072,0.006577777779764599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,3584,0.08816444211535984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,3584,0.050717333952585854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,3072,0.04320177767011854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,2560,0.005944889038801193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,3072,0.07894577582677205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,2048,0.0053022222386466126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,2560,0.03601244423124526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,2560,0.06863555643293592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,1536,0.004920888692140579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,2048,0.030951112508773804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,2048,0.0565662251578437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,1024,0.0041751112374994485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,1536,0.04801866743299696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,1536,0.025812443759706285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,1024,0.01995644470055898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,768,0.003930666794379552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,6144,0.009915555516878763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,1024,0.03744622071584066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,512,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,768,0.01759999990463257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,768,0.033928887711630926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,256,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,512,0.01475822263293796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,512,0.028816001282797918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,128,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,256,0.012422222230169507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,256,0.024732444021436904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,64,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,128,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,128,32,0.0030817778574095834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,2048,128,128,0.02164799968401591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,65536,0.04821422364976671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,64,0.01144177797767851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,128,32,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,51200,0.040044443474875555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,16384,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,16384,0.24193067020840117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,12288,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,51200,0.72781064775255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,10240,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,12288,0.18333421813117134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,65536,0.927559110853407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,8192,0.009701333112186855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,7168,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,10240,0.1521715588039822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,8192,0.11857867240905762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,6144,0.008354666332403818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,7168,0.10127111275990804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,5120,0.007954667011896769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,4096,0.007644444704055786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,5120,0.0720506641599867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,6144,0.08578399817148845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,3584,0.0069520001610120135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,3072,0.006281777802440856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,3584,0.05067911081843906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,2560,0.005879111174080107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,3072,0.04346933298640781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,2048,0.005168000029193031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,2560,0.03771022293302748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,1536,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,2048,0.030692444907294378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,1024,0.003957333250178231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,1536,0.02537333303027683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,768,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,1024,0.020263999700546265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,512,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,768,0.017567111386193168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,256,0.003167111012670729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,512,0.014538667268223233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,128,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,256,0.012404444317022959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,64,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,4096,0.05819911426968045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,128,0.011648000114493899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,64,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,65536,0.04656622144911024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,64,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,64,32,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,51200,0.03886311252911886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,16384,0.00943377779589759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,16384,0.24172888861762154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,12288,0.00850311087237464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,51200,0.7287182278103299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,10240,0.007989333735571967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,65536,0.9275439580281576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,12288,0.18301867114173043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,8192,0.012028444144460889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,7168,0.011035555766688453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,10240,0.15279022852579752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,8192,0.11836355262332493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,6144,0.010010666317409938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,7168,0.10134755240546332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,5120,0.009358222285906473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,4096,0.008338666624493068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,6144,0.08594133456548055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,3584,0.006903111106819577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,5120,0.07221955723232694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,4096,0.05865866608089871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,3072,0.006240889016124938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,2560,0.0057760001056724125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,3584,0.050670223103629224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,2048,0.005220444252093633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,3072,0.04340799980693393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,2560,0.03734666771358914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,1536,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,1024,0.0038559999730851916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,2048,0.030547555949952867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,1536,0.02544266647762722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,768,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,1024,0.020278222031063504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,768,0.01756800048881107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,512,0.014491554763582019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,256,0.012157333393891653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,128,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,128,0.01145066652033064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,64,0.0026026666164398193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,512,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,32,0.0027031112048361036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2048,32,256,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,64,0.011704000333944956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2048,32,32,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,16384,1.9650692409939234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,16384,2.6124560038248696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,51200,5.976593017578125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,12288,1.474930657280816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,10240,1.2097661760118272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,12288,1.8205430772569444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,8192,0.9648542404174805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,51200,8.892940097384983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,10240,1.4900959862603083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,8192,1.179310268825955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,7168,0.8126426802741157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,6144,0.7173022164238824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,7168,1.0143333011203342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,6144,0.8618533346388074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,5120,0.6891137758890787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,4096,0.4644062254163954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,5120,0.715103997124566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,3584,0.42778576744927305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,4096,0.5710479948255751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,3072,0.3666977882385254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,2560,0.2981022198994954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,3584,0.5013342433505589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,3072,0.45000267028808594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,2048,0.23794489436679414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,1536,0.18058844407399496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,2560,0.379854228761461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,1024,0.13211999999152288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,768,0.0988968875673082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,1536,0.23988710509406197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,2048,0.302440881729126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,1024,0.1811537742614746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,512,0.07662044631110297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,256,0.0635768903626336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,128,0.051318221622043185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,768,0.1461271047592163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,256,0.11123377747005886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,64,0.04982666505707634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,512,0.12480978171030681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,65536,32,0.0529075562953949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,128,0.10610755284627278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,64,0.1059839990403917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,65536,32,0.10749333434634739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,51200,4.479621463351779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,65536,5.715710110134548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,16384,1.507515589396159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,12288,1.1072364383273654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,51200,7.058180491129558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,16384,2.056157430013021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,10240,0.9790933397081164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,65536,8.922574361165365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,12288,1.4503084818522136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,8192,0.7864391009012858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,10240,1.1923546261257596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,7168,0.7099644343058268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,6144,0.6036062240600586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,8192,0.9268435372246636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,7168,0.8108168707953559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,5120,0.4974817699856228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,6144,0.6903688642713758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,5120,0.5721626811557347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,4096,0.3804577721489801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,3584,0.33545600043402773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,4096,0.47420533498128253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,3584,0.4005911085340712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,3072,0.2918311225043403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,2560,0.23175554805331758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,3072,0.34646044837103945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,2048,0.2028648853302002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,1536,0.1487191120783488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,2560,0.295197327931722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,1024,0.11651021904415554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,2048,0.24270844459533691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,768,0.09274222453435262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,1024,0.14660622013939753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,512,0.0704417758517795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,1536,0.19361510541703966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,768,0.11865688694847955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,256,0.05005955696105957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,128,0.03887822230656942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,64,0.033923556407292686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,512,0.1008595559332106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,51200,32,0.035749332772360906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,128,0.08570488956239487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,256,0.09060800075531006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,64,0.08586221933364868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,51200,32,0.08665600087907578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,51200,1.5519866943359375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,65536,2.001396391126845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,16384,0.5113555590311686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,12288,0.38788265652126735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,51200,2.5930906931559243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,16384,0.7617226706610786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,10240,0.3001848856608073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,12288,0.5538258022732204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,10240,0.4543324576483832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,8192,0.24226488007439506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,65536,3.3222836388481984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,7168,0.20993244647979736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,8192,0.35780800713433164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,6144,0.18276000022888184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,5120,0.15394311481051975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,7168,0.3138817681206597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,6144,0.26829777823554146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,4096,0.12343645095825195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,5120,0.2242835627661811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,3584,0.10690666569603814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,4096,0.18072977330949572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,3072,0.0928071141242981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,2560,0.07875555753707886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,3584,0.1603271166483561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,3072,0.14036444822947183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,2048,0.06492800182766385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,1536,0.050423112180497914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,2048,0.10012266370985244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,2560,0.12116177876790364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,1536,0.08123555448320177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,768,0.029130667448043823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,768,0.05058222346835666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,1024,0.062063111199273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,512,0.02102577851878272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,256,0.01829777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,256,0.035664889547559954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,512,0.04155377878083123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,128,0.014821334017647637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,64,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,1024,0.0367253323396047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,128,0.034154666794670954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,64,0.03362311257256402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,16384,32,0.013189333180586496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,16384,32,0.03335466649797227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,65536,1.4824861950344508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,51200,1.1701840294731987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,16384,0.39519111315409344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,65536,2.6927341885036893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,16384,0.6116693284776475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,12288,0.2932026651170519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,10240,0.24392000834147134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,12288,0.4490728908114963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,51200,2.0803777906629777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,8192,0.1990977790620592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,10240,0.37081334326002335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,7168,0.17548622025383842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,8192,0.2903742260403103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,6144,0.1526871124903361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,7168,0.2540115515391032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,5120,0.12970488601260716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,6144,0.21876977549658883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,4096,0.11892267068227132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,5120,0.18374133110046387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,3584,0.10467555787828232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,3072,0.08851910961998834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,3584,0.13221600320604113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,2560,0.0747893320189582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,2048,0.0598942240079244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,2560,0.09973600175645615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,3072,0.1162702242533366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,1536,0.04754844307899475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,1024,0.0343466666009691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,2048,0.08315733406278822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,768,0.026757333013746474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,1536,0.06766488817003039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,512,0.019527110788557265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,768,0.0413333343134986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,256,0.014450665977266101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,256,0.02886311213175456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,512,0.03501066565513611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,128,0.014208889669842191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,64,0.012067555553383298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,128,0.02827733423974779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,12288,32,0.012847999731699625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,1024,0.05063466562165154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,32,0.027830223242441814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,4096,0.1490239964591132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,12288,64,0.027810666296217177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,51200,0.9602337943183051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,65536,1.2256124284532335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,16384,0.3173484537336561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,12288,0.23697156376308867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,16384,0.5444266531202528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,10240,0.19719821876949736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,12288,0.3975502120123969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,65536,2.33622317843967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,10240,0.328853342268202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,8192,0.16714578204684785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,51200,1.835874769422743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,7168,0.13948267036014134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,6144,0.12756889396243626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,7168,0.22522489229838052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,5120,0.10398933622572158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,8192,0.25851554340786403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,4096,0.08582044310039944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,3584,0.07575644387139215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,5120,0.16253244876861572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,3072,0.06335288948482938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,4096,0.13372355037265354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,3584,0.11795732710096572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,2560,0.05347822109858195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,3072,0.10404266913731892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,2048,0.044023109806908496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,2560,0.08948266506195068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,1536,0.03487822082307603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,2048,0.07447289095984565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,1024,0.024767110745112102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,1536,0.05990844302707248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,6144,0.19400355550977919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,768,0.020243555307388306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,512,0.01568000018596649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,768,0.037471999724706016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,256,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,512,0.030100444952646892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,128,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,256,0.026224889689021643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,64,0.010022222167915767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,128,0.025070221887694463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,10240,32,0.010078222387366826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,64,0.025026667449209426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,32,0.024711999628278945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,10240,1024,0.04450222187572055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,65536,0.991795539855957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,51200,0.7777742279900445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,16384,0.2568506664699978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,16384,0.47204266654120547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,65536,2.0466461181640625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,12288,0.19568088319566515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,10240,0.164192001024882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,12288,0.34786399205525714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,51200,1.5793155034383137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,8192,0.13473688231574163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,7168,0.11914843983120388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,10240,0.2866720093621148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,8192,0.22511998812357584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,6144,0.10307733217875163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,5120,0.08774755398432414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,7168,0.1969511111577352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,6144,0.16927466127607557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,4096,0.07178399960199992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,3584,0.06295110781987508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,5120,0.14230488406287298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,4096,0.11592443784077962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,3072,0.055007110039393105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,3584,0.10425955719417995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,2560,0.04567377765973409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,2048,0.037556442949506975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,3072,0.09202222029368083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,1536,0.029871109459135268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,2560,0.07895021968417697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,1024,0.022450667288568284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,2048,0.06562399864196777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,768,0.017913776967260573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,1536,0.05273422267701891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,1024,0.039423998859193586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,768,0.033944000800450645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,256,0.010020444790522257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,512,0.02648711038960351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,128,0.008601778083377415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,256,0.023352000448438857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,64,0.007368000017272101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,128,0.02235200007756551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,32,0.007616000043021307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,64,0.02258488867017958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,8192,32,0.02197333342499203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,8192,512,0.013791110780504016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,65536,0.8183431095547147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,16384,0.2188133398691813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,51200,0.6406399938795302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,12288,0.16612177424960667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,16384,0.43477333916558164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,10240,0.1385893291897244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,12288,0.3239786624908447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,65536,1.8861103057861328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,51200,1.443312856886122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,10240,0.2663324409061008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,7168,0.10057599676979913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,8192,0.2087706724802653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,6144,0.08354222112231785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,7168,0.1822648843129476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,5120,0.07005866368611653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,6144,0.1567644410663181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,4096,0.06151022513707479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,5120,0.13212266233232287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,4096,0.10754844215181138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,3072,0.050123555792702564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,3584,0.09609511163499619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,2560,0.03749600052833557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,8192,0.11109333568149143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,2048,0.031153778235117596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,3072,0.08555289109547932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,1536,0.03050755461057027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,2560,0.07377066877153185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,1024,0.0181013329161538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,2048,0.060841778914133705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,768,0.017045333981513977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,1536,0.048764444059795804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,512,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,1024,0.03705155518319871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,768,0.03191555539766947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,256,0.009087111386987898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,512,0.02537155482504103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,128,0.008640888664457533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,3584,0.05281866590181986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,128,0.020968889196713764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,64,0.009168888959619734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,7168,32,0.007968000239796108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,64,0.020999110407299463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,32,0.021184000704023574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,65536,0.6989910867479113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,7168,256,0.02197066611713833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,51200,0.5463786655002171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,16384,0.19036266538831922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,16384,0.41335556242201066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,51200,1.3344969219631617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,65536,1.6834968990749781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,12288,0.1433146662182278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,10240,0.12464800145890977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,8192,0.09591199954350789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,10240,0.2448444366455078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,12288,0.29898044798109263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,7168,0.08621333042780559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,6144,0.07366933425267537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,8192,0.1936337815390693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,5120,0.06146844228108724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,7168,0.16856443881988525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,6144,0.1447395616107517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,4096,0.05144266618622673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,3584,0.04623644550641378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,4096,0.09982755449083115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,5120,0.12273511621687148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,3072,0.03887733485963609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,2560,0.03477777706252204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,3584,0.08918399943245782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,2048,0.027671110298898485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,3072,0.07793777518802218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,2560,0.06810666455162896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,1536,0.022750222020679053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,1024,0.016664889123704698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,1536,0.044864889648225575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,2048,0.05662755833731758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,768,0.013190222283204397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,512,0.010843555960390302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,1024,0.03458311160405477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,256,0.010058666268984476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,512,0.023337778117921617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,768,0.02961155441072252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,128,0.007312888900438945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,64,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,256,0.020837333467271592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,6144,32,0.006722666737106111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,64,0.019567999574873183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,128,0.020041777027977836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,6144,32,0.0196115556690428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,65536,0.8055822054545084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,51200,0.6322435802883571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,16384,0.2077102263768514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,65536,1.5408817927042644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,16384,0.3672213289472792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,12288,0.15707111358642578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,10240,0.1052817768520779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,51200,1.2048924763997395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,8192,0.08629866441090901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,12288,0.2752773231930203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,7168,0.07168533404668172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,10240,0.22649065653483072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,8192,0.17608177661895752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,6144,0.06363644202550252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,5120,0.0566435522503323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,7168,0.1552542183134291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,6144,0.13289422459072536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,4096,0.042436444097095065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,3584,0.03989244500796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,3072,0.03355377912521362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,5120,0.1130524476369222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,4096,0.09274755583869086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,2560,0.028397334946526423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,3584,0.08216000265545316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,2048,0.023825777901543513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,3072,0.0709022217326694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,1536,0.019846222466892667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,2560,0.06059377723269992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,1024,0.014811555544535318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,1536,0.041857779026031494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,768,0.012494222157531314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,2048,0.05146844519509209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,512,0.010442666709423065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,1024,0.03198933270242479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,256,0.008006222546100616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,128,0.007308444215191736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,768,0.026833777626355488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,64,0.006590222318967183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,256,0.01921688848071628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,512,0.02201333310869005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,5120,32,0.006595555692911148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,64,0.018256000346607633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,128,0.01854311095343696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,5120,32,0.01853155593077342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,65536,0.47847286860148114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,51200,0.37326934602525497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,16384,0.12763822078704834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,16384,0.3350257873535156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,51200,1.0761679543389213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,12288,0.09869244363572861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,65536,1.3902817832099066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,10240,0.08114844560623169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,8192,0.06659555435180664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,10240,0.20558133390214708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,12288,0.25127466519673664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,7168,0.0586986674202813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,6144,0.051184889343049794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,8192,0.16167822149064806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,5120,0.04402133491304186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,7168,0.1417671177122328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,6144,0.12168178293440078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,4096,0.03459466828240289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,3584,0.03108355402946472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,4096,0.08415555291705662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,5120,0.10327288839552139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,3072,0.026717333330048457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,2560,0.023391111029518977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,3584,0.0750382211473253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,2048,0.01999022232161628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,3072,0.06657333506478204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,2560,0.05585600270165337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,1536,0.016695999436908297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,1024,0.01257333325015174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,1536,0.03792533278465271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,2048,0.04628622200753954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,768,0.010144000252087912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,512,0.008046222229798635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,1024,0.028394665982988145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,256,0.006380444599522485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,512,0.02032355633046892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,768,0.026027555267016094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,128,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,64,0.0052355556852287715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,256,0.017634666628307767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,128,0.01756444407833947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,64,0.01687022215790219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,4096,32,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,4096,32,0.017251556118329365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,65536,0.42041956053839785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,51200,0.32683467864990234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,16384,0.11083733373218113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,65536,1.2930079566107857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,16384,0.31936976644727916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,51200,1.0179689195421007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,12288,0.08341155449549358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,10240,0.07703377803166707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,8192,0.058021333482530385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,12288,0.23924978574117026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,10240,0.19642133182949492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,7168,0.05089599887530009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,8192,0.1532720062467787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,6144,0.04381600022315979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,5120,0.03748800026045905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,6144,0.1157004435857137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,7168,0.1341164509455363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,4096,0.031896889209747314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,3584,0.028392000330819026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,5120,0.09805066717995538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,3072,0.02475555572244856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,4096,0.08073333236906263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,2560,0.02184355590078566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,3072,0.06255022022459242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,2048,0.018072888255119324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,2560,0.05279911226696438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,1536,0.014775110615624322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,2048,0.043806221750047475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,1024,0.010799110763602786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,1536,0.036157333188586764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,768,0.009120889008045197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,1024,0.02715911136733161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,512,0.007415999968846639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,3584,0.07139644357893202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,256,0.006325333482689328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,768,0.023789332972632513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,128,0.005972444597217772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,512,0.019631110959582858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,256,0.017172444197866652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,64,0.00564533347884814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3584,32,0.005720888988839255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,128,0.01661066710948944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,64,0.016216889023780823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3584,32,0.016374222106403775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,65536,0.3409430980682373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,51200,0.2713377740648058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,16384,0.09400622049967448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,12288,0.07235199875301786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,16384,0.3029537730746799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,10240,0.06407022476196289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,51200,0.9590959548950195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,65536,1.2312488555908203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,8192,0.0522328880098131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,7168,0.04324177900950114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,10240,0.1871217754152086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,8192,0.1462799972958035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,6144,0.03797600004408095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,7168,0.1283244424396091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,5120,0.03199022346072727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,6144,0.11022133297390407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,4096,0.027779555983013574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,5120,0.09409333599938287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,3584,0.02384888960255517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,4096,0.07709422376420763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,3072,0.021935110290845234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,12288,0.22774844699435762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,2560,0.018751111295488145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,3584,0.06909422079722087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,2048,0.01569600072171953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,3072,0.05888266695870293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,1536,0.012815111213260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,2560,0.05114666620890299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,1536,0.03428533342149522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,1024,0.025772444076008264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,768,0.00870666652917862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,768,0.022682666778564453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,512,0.0070053330726093715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,512,0.018627555833922494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,256,0.005701333284378052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,2048,0.041815999481413096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,256,0.016231111354298063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,1024,0.009782222410043081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,128,0.00489066665371259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,64,0.004947555561860402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,3072,32,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,128,0.015899555550681222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,64,0.015575110912322998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,3072,32,0.015555555621782938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,65536,0.3188675509558784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,51200,0.24835199779934355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,16384,0.08599644237094456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,65536,1.1388613382975261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,16384,0.2862079938252767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,51200,0.8972142537434896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,12288,0.06691910823186238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,10240,0.05852444304360283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,8192,0.04634488953484429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,10240,0.17695644166734484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,12288,0.2153937816619873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,7168,0.04044444362322489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,8192,0.13824089368184408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,6144,0.03391199972894456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,5120,0.02851822310023838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,7168,0.12051999568939209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,6144,0.10393689076105754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,5120,0.08817866775724624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,3584,0.0210506667693456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,4096,0.0722915530204773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,3072,0.019288889235920377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,3584,0.06436266501744588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,2560,0.016578666037983365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,3072,0.054970665110482105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,2048,0.013651555611027611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,2560,0.047263112333085805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,1536,0.010784000158309937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,2048,0.039156443542904325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,1024,0.0088755554623074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,1536,0.03221688999070062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,768,0.007676444119877285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,1024,0.02442311081621382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,4096,0.02366488840844896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,512,0.006648889018429651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,256,0.005606222069925732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,768,0.021967111362351313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,128,0.004925333377387789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,512,0.017988445030318368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,256,0.015618667006492615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,64,0.00462400002612008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2560,32,0.004600888739029567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,64,0.014855111638704935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,32,0.01484444406297472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,65536,0.25973688231574166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,51200,0.2071839968363444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2560,128,0.015180443723996481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,16384,0.06854844755596586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,65536,1.0809982087877061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,51200,0.847692436642117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,12288,0.053584890233145825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,16384,0.27203822135925293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,10240,0.04660444458325704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,8192,0.03525955478350321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,12288,0.20457333988613555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,7168,0.030497776137457952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,10240,0.1689457760916816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,6144,0.025632888078689575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,8192,0.13203822241889104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,7168,0.1148408916261461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,5120,0.022504000200165644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,6144,0.09946578078799778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,4096,0.01901333365175459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,3584,0.017255999975734286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,5120,0.08374577760696411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,3072,0.01520088811715444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,4096,0.06878133614857991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,2560,0.01332622187005149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,3072,0.052689777480231396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,3584,0.06083910995059543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,2048,0.011630222201347351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,2560,0.04553777641720242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,1536,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,2048,0.036914666493733726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,1024,0.007679111427730984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,1536,0.03122666809293959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,768,0.006639110959238476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,1024,0.023440000083711412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,512,0.005893333504597346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,768,0.0209840006298489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,256,0.004886222382386525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,512,0.01718933383623759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,128,0.004274666723277834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,256,0.01514933341079288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,64,0.00423911131090588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,128,0.014461333552996317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,2048,32,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,64,0.014324444863531323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,2048,32,0.014140443669425117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,65536,0.18757332695855033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,51200,0.14648266633351645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,65536,0.9878426657782661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,51200,0.7767386436462402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,16384,0.05102488729688856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,12288,0.03958844476275974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,16384,0.25447199079725474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,10240,0.03198488884501987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,12288,0.19135733445485434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,8192,0.028212444649802312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,10240,0.157331559393141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,8192,0.122670226626926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,7168,0.10690844721264309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,6144,0.021543110410372417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,6144,0.09221866395738389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,5120,0.018274666534529794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,5120,0.07758488919999865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,4096,0.015434667468070984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,7168,0.023384888966878254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,3584,0.01404355631934272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,4096,0.06304088566038343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,3072,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,3584,0.055551111698150635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,2560,0.010727110836240979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,3072,0.048264887597825795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,2048,0.009390222529570261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,2560,0.04210844304826525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,1536,0.008115555677149031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,2048,0.03389422098795573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,1024,0.006996444529957241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,1536,0.028773334291246202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,768,0.006210666563775804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,1024,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,512,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,768,0.01959288948112064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,256,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,512,0.016193777322769165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,128,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,256,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,64,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,128,0.013767110804716745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1536,32,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,64,0.013541333377361298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1536,32,0.013438222308953604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,65536,0.15232355064815944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,51200,0.12033066484663223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,65536,0.9375279744466146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,51200,0.7349555757310656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,16384,0.0409840014245775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,12288,0.031089779403474595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,10240,0.024881778491867915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,16384,0.24255111482408312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,8192,0.01925155520439148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,10240,0.14986666043599448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,7168,0.01685599982738495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,12288,0.18318577607472739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,8192,0.11752800146738689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,6144,0.015021334091822306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,7168,0.10154133372836643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,5120,0.013129777378506131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,4096,0.011672889192899069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,6144,0.08734577894210815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,3584,0.010204444328943888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,5120,0.07386310895284016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,3072,0.008930666579140557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,2560,0.008195555872387355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,3584,0.05246399839719137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,2048,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,4096,0.060115555922190346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,3072,0.04479111234347025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,1536,0.006370666540331311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,1024,0.005423111220200856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,2048,0.032487110959159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,1536,0.027202667461501226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,2560,0.03909688856866624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,768,0.004615111069546806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,512,0.004354666504595014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,256,0.003936888857020272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,768,0.018568888306617737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,512,0.015789333317014907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,128,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,128,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,256,0.014293332894643148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,64,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,1024,32,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,32,0.012798222402731577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,64,0.013188444077968597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,65536,0.12221777439117432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,51200,0.09631822506586711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,1024,1024,0.02107200026512146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,16384,0.03439822130733066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,65536,0.8993075688680013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,51200,0.7055306434631348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,12288,0.025782222549120586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,16384,0.23416267500983345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,10240,0.022263111339675054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,8192,0.018703111343913607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,12288,0.1757297780778673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,7168,0.016179554992251925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,10240,0.145179549853007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,6144,0.01516266663869222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,7168,0.09772977564069961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,5120,0.012536889149083031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,6144,0.08365777466032241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,5120,0.07079555590947469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,4096,0.011216888825098673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,3584,0.01035911093155543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,4096,0.056847108734978564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,3072,0.008671110702885522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,3584,0.0499875545501709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,2560,0.008392000363932716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,8192,0.11318311426374648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,2048,0.007346666521496243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,2560,0.03682488865322537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,1536,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,2048,0.03125244379043579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,1024,0.005599110904667113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,1536,0.02624622152911292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,768,0.004905777672926585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,1024,0.020584000481499564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,512,0.004304000073009067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,3072,0.04388622111744351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,768,0.018340445227093168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,256,0.0036195553839206696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,128,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,512,0.015607110328144498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,64,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,256,0.013167999684810638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,128,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,768,32,0.003313777761326896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,64,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,65536,0.08737422360314263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,768,32,0.012400888734393649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,51200,0.06986844539642334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,16384,0.027263111538357202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,16384,0.22447466850280762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,51200,0.674350208706326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,12288,0.02118844456142849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,10240,0.025814221964942083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,65536,0.858966244591607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,8192,0.020984888076782227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,12288,0.16865244176652697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,7168,0.019117333822780185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,10240,0.13966577582889134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,6144,0.017265778448846605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,8192,0.10848533444934422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,7168,0.09339466359880234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,5120,0.013820444544156393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,4096,0.008649778034951951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,5120,0.06771200233035617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,6144,0.08019466532601251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,3584,0.007696888513035244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,3072,0.007347555624114142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,4096,0.05442044470045301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,2560,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,3584,0.047359112236234874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,3072,0.041610668102900185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,2048,0.006008000009589725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,1536,0.0052888890107472735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,2560,0.0357893341117435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,2048,0.02974933385848999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,1024,0.004918222212129169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,768,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,1024,0.019664888580640156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,1536,0.025069332785076563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,512,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,256,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,768,0.017940445078743827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,128,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,512,0.01498311095767551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,64,0.0031840000301599503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,128,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,512,32,0.003257777748836411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,65536,0.05907999806933933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,64,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,51200,0.04827466607093811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,32,0.012033777932325998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,16384,0.018555555078718398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,16384,0.2143333355585734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,12288,0.016492444607946608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,512,256,0.01274577776590983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,51200,0.6436755392286513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,10240,0.014018666413095264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,12288,0.16151289145151773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,8192,0.012337778177526263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,10240,0.13298132684495714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,8192,0.10259644190470378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,7168,0.0886240005493164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,6144,0.00996088898844189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,65536,0.8200071122911242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,6144,0.076310224003262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,5120,0.008991999758614434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,4096,0.007667555577225155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,5120,0.06440977917777167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,3584,0.0070364442136552595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,4096,0.051049778858820595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,7168,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,3072,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,3584,0.045328888628217906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,3072,0.03928622272279527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,2560,0.033991999096340604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,2048,0.005631999837027655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,1536,0.00498577786816491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,2048,0.028175999720891316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,1024,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,1536,0.024116444918844435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,1024,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,768,0.016921778519948322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,2560,0.006036444256703059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,512,0.014547554983033074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,256,0.003350222276316749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,256,0.012433778080675336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,128,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,768,0.003986666599909465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,64,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,128,0.011949333051840464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,32,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,256,512,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,65536,0.04788622260093689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,64,0.01146311147345437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,51200,0.03997777899106344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,256,32,0.011828444070286222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,16384,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,16384,0.21450932820638022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,12288,0.010008888940016428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,65536,0.8189599778917102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,51200,0.6429297659132216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,10240,0.009006222089131674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,8192,0.0103502223889033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,12288,0.16165333324008518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,10240,0.13274133205413818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,7168,0.00979555563794242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,8192,0.10276355346043904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,6144,0.010046222143703038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,5120,0.00869422240389718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,6144,0.07634933127297296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,4096,0.007348444726732042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,5120,0.06444088617960612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,3584,0.006963555183675554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,4096,0.05072177780999077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,3072,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,3072,0.03834755553139581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,3584,0.04491466614935133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,2560,0.005609777652555042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,2048,0.00518133325709237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,2560,0.03330755564901564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,2048,0.02838311261600918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,1536,0.004896000027656555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,1024,0.00425866660144594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,1536,0.024158222807778254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,7168,0.08831377824147542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,768,0.003929777691761653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,512,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,1024,0.01924977699915568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,256,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,512,0.014148443937301636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,768,0.016944888565275405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,128,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,256,0.012451555165979596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,128,32,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,64,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,128,0.011810666984981961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,65536,0.04199999901983473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,51200,0.03498399919933743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,128,32,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,16384,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,12288,0.00978311151266098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,16384,0.21435022354125977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,10240,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,51200,0.6425466537475586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,65536,0.8191653357611762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,12288,0.1608897712495592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,7168,0.009321778184837764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,10240,0.13178755177391901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,8192,0.10233510865105523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,6144,0.00869333330127928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,7168,0.08819022443559434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,5120,0.008362666600280339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,4096,0.007339555356237624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,5120,0.06401866674423218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,6144,0.07563733392291598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,3584,0.00699733363257514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,4096,0.05085066623157925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,3072,0.0063271112740039825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,3584,0.04440800017780728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,8192,0.008277333445019191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,3072,0.03843733337190416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,2560,0.005660444498062134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,2048,0.005251555393139522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,1536,0.00461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,2560,0.03291733397377862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,1024,0.003963555726740095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,2048,0.02812088860405816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,768,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,1536,0.02438755498992072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,1024,0.018978666928079393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,512,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,256,0.0033173333439562055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,512,0.01423555612564087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,768,0.016933333542611863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,64,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,256,0.012177777787049612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,64,32,0.002967111145456632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,64,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,65536,0.040163556734720864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,32,0.011682666838169098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,51200,0.03341333402527703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,65536,0.8184577624003092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,16384,0.009733333355850643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,51200,0.6429182158576118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,12288,0.013215111361609565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,16384,0.21398844983842638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,10240,0.012799999780125089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,12288,0.16145955191718206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,8192,0.010442666709423065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,10240,0.1321022245619032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,7168,0.016230222251680162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,8192,0.10250666406419541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,7168,0.08810133404201931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,6144,0.010036444498433007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,5120,0.009724444813198512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,64,128,0.011496889094511667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,6144,0.07578311363855998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,4096,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,3584,0.008372444245550368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,5120,0.06405511167314318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,3072,0.006981333096822103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,4096,0.050740443997912936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,2560,0.006731555693679386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,3584,0.04430222180154589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,2048,0.005992888990375731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,3072,0.0380968881977929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,2560,0.032829334338506065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,1536,0.004325333154863781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,1024,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,1536,0.02407022151682112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,768,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,2048,0.028247998820410833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,512,0.003293333368168937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,1024,0.01904533306757609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,768,0.017109332813156974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,256,0.0032702222880389956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,128,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,256,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,512,0.014262222581439547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,64,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1793,32,32,0.0028204443968004654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,128,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,64,0.011202666494581433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1793,32,32,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,16384,1.7391839557223852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,51200,5.358033921983506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,16384,2.5208577050103083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,12288,1.3106444676717122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,51200,8.59140099419488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,16384,1.8556427425808375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,12288,1.7715671327379015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,10240,1.2153084013197157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,51200,5.933652242024739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,12288,1.4070213105943468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,10240,1.44342221154107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,8192,0.9684097501966688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,10240,1.1775964101155598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,8192,1.1344950993855794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,7168,0.8474666807386609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,8192,0.9437253740098742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,7168,0.9750604629516602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,6144,0.7194079822964139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,5120,0.6185262468126085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,6144,0.7212150891621908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,5120,0.7076240115695529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,7168,0.8324897554185656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,4096,0.46518580118815106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,5120,0.6126391092936198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,4096,0.5506053500705296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,3584,0.4266968833075629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,4096,0.5008417765299479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,6144,0.8313155704074435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,3584,0.48359023200141055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,3072,0.3402862283918593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,3072,0.41884178585476345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,3584,0.4456782341003418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,2560,0.2906400097741021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,3072,0.39177510473463273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,2048,0.2328284316592746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,2560,0.35855287975735134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,2560,0.3352479934692383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,1536,0.1923546658621894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,2048,0.28215734163920086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,1536,0.23378578821818033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,1024,0.12632266680399576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,1536,0.2276782194773356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,1024,0.17733689149220785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,768,0.0935644441180759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,2048,0.2949511210123698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,768,0.14297244283888075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,768,0.1474133332570394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,512,0.12128622002071804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,512,0.11985333760579427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,256,0.055163555675082736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,1024,0.17518756124708387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,256,0.10839200019836426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,128,0.047150221135881215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,512,0.07143288850784302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,256,0.09745866722530788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,64,0.03989955451753404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,128,0.10294489065806071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,65536,128,0.0772835546069675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,65536,32,0.041407111618253924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,64,0.10383288727866279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,65536,32,0.10424444410536025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,65536,5.495363447401259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,51200,4.316864861382379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,65536,8.599146525065104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,16384,1.5177182091606989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,65536,6.084509531656901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,51200,6.8000060187445746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,16384,2.0062639448377824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,12288,1.0424577924940321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,12288,1.4097973505655925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,16384,1.5094310972425673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,10240,0.997247060139974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,10240,1.1502035988701713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,12288,1.1411387125651042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,8192,0.7453457514444987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,8192,0.9092195298936633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,10240,0.9511440065171984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,51200,4.7915831671820746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,7168,0.6460728645324707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,8192,0.7665208710564507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,6144,0.5563831329345703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,7168,0.7802871068318685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,5120,0.48552799224853516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,7168,0.677912023332384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,6144,0.5860728687710232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,5120,0.5577724244859483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,4096,0.36199198828803164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,5120,0.49785598119099933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,6144,0.6638195779588487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,3584,0.34126665857103133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,4096,0.44420356220669216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,4096,0.4087217648824056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,3584,0.3872906631893582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,3072,0.2738115522596571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,2560,0.23002934455871582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,3584,0.36293866899278426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,3072,0.3189093271891276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,2560,0.28888000382317436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,2048,0.18281243907080758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,2560,0.2745680014292399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,1536,0.13845245043436685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,2048,0.2300257682800293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,2048,0.2381626764933268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,3072,0.3397973378499349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,1024,0.094541331132253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,1536,0.18944444921281603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,768,0.07499733236100939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,1536,0.18627911143832734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,1024,0.14300266901652017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,512,0.058527999454074435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,1024,0.14447111553615993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,768,0.12212177117665608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,256,0.04385333259900411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,512,0.09835022025638157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,512,0.10028266906738281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,256,0.08807200193405151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,128,0.04122844338417053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,256,0.08074133263693915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,64,0.03542755709754096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,128,0.08363644282023112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,51200,32,0.03695644603835212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,51200,128,0.06665155622694227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,64,0.08346044354968601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,32,0.08440088563495213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,51200,768,0.11630666255950928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,65536,1.7555661731296117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,51200,1.2971910900539823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,65536,3.260270224677192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,16384,0.42045066091749406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,51200,2.521699481540256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,65536,2.676515579223633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,51200,2.1012969546847873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,16384,0.7402426401774088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,12288,0.3162924448649089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,16384,0.6875031259324816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,10240,0.297727108001709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,12288,0.5351582103305393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,10240,0.4412026670244005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,12288,0.5240302085876465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,10240,0.43591822518242734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,8192,0.2378177907731798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,7168,0.20930400159623888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,8192,0.3463199933369954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,8192,0.34962044821845156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,7168,0.330039103825887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,6144,0.1797413296169705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,7168,0.30817333857218426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,6144,0.2610035472446018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,5120,0.15004177888234457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,6144,0.2675973309410943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,5120,0.21836088763342965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,4096,0.12026845084296332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,5120,0.2282177872127957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,4096,0.1769146654340956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,3584,0.10810222228368123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,4096,0.1892320050133599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,3072,0.09001511335372925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,3584,0.15680444240570068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,3584,0.16885777314503989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,2560,0.07565689086914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,3072,0.14895110660129124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,3072,0.1378906700346205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,2048,0.06444622410668267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,2560,0.12952621777852377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,2560,0.11862756146325006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,1536,0.04635555545488993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,2048,0.11034133036931355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,2048,0.09879200326071845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,1024,0.033019555939568415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,1536,0.0796462231212192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,1536,0.09100977579752605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,768,0.026682666606373254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,1024,0.06092355648676554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,1024,0.07204977671305339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,768,0.049343112442228526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,512,0.020647111866209242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,768,0.06316444608900282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,512,0.04041066765785217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,256,0.016386666231685214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,512,0.05247733328077528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,128,0.014458666245142618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,256,0.03533422284656101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,256,0.044734223021401294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,16384,128,0.03629422187805176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,64,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,128,0.033144887950685285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,16384,32,0.012479110724396177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,64,0.033196445968416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,16384,32,0.032285332679748535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,65536,1.391982184516059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,51200,1.038903130425347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,65536,2.5751724243164062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,16384,0.3249582184685601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,51200,2.0251021915011935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,16384,0.5951449076334635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,65536,2.291012446085612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,12288,0.23785866631401908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,12288,0.4372488922542996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,16384,0.5916204452514648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,10240,0.2464524375067817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,12288,0.44813956154717344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,8192,0.18635822667015922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,10240,0.36017778184678817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,10240,0.3747226662105984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,8192,0.2827315595414904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,7168,0.16406133439805773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,8192,0.30011023415459526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,7168,0.24681332376268175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,6144,0.14075199762980142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,7168,0.2646568881140815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,6144,0.21260889371236166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,51200,1.8004667494032118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,5120,0.12091022067599827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,6144,0.23013154665629068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,4096,0.09757955869038899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,5120,0.17959466245439318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,5120,0.19611822234259713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,3584,0.0842462248272366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,4096,0.16218488746219212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,3584,0.12979022661844888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,3072,0.0741128921508789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,3584,0.1446346706814236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,2560,0.06292355722851224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,3072,0.11415910720825195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,3072,0.1290337774488661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,2560,0.0985528892940945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,2048,0.052319112751219005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,2560,0.11251821782853867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,1536,0.040391110711627536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,2048,0.08173600170347425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,2048,0.0964391099082099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,1536,0.06598399745093451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,1024,0.0284222231970893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,4096,0.14671199851565891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,1536,0.07978933387332492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,768,0.023434667123688593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,1024,0.049673779143227466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,1024,0.06368355618582831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,512,0.01811911165714264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,768,0.05604177713394165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,768,0.04083555605676439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,256,0.013743110828929476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,512,0.04661688870853848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,512,0.033187554942237005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,128,0.011052444577217102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,256,0.03966044386227926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,256,0.02852355440457662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,64,0.009836444424258338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,12288,128,0.03163377775086297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,128,0.027466666367318895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,12288,32,0.009729777773221334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,64,0.027281777726279363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,12288,32,0.027151111099455092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,65536,1.1107653511895073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,51200,0.8709031211005317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,65536,2.293840832180447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,65536,2.157828436957465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,51200,1.7727022171020508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,16384,0.2987119886610243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,16384,0.5277751286824545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,12288,0.2125048902299669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,16384,0.5588346587287055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,51200,1.6943848927815754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,12288,0.3881591161092122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,10240,0.2002293268839518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,12288,0.4239635467529297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,10240,0.31949599583943683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,8192,0.16272621684604222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,10240,0.3527377711402045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,8192,0.25143644544813365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,7168,0.14313777287801108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,8192,0.2827297846476237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,6144,0.1233359972635905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,7168,0.21959733963012695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,7168,0.24900266859266493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,5120,0.09550311168034871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,6144,0.19018756018744576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,6144,0.21591556072235107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,5120,0.1836320029364692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,4096,0.08067644304699369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,5120,0.15892444716559515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,3584,0.07109510898590088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,4096,0.13049066066741943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,4096,0.15291555722554526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,3072,0.06242933538224962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,3584,0.11662133534749348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,3584,0.13770043849945068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,2560,0.05164533191257053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,3072,0.10245866907967462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,3072,0.1215475532743666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,2048,0.041952000723944775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,2048,0.07342133257124159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,2560,0.08869777785407172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,2048,0.09091111024220784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,1536,0.03296800123320685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,1024,0.02385866641998291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,1536,0.07533155547247992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,1024,0.04390133420626322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,1536,0.059049778514438205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,768,0.01954844428433312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,1024,0.0598906675974528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,512,0.014666666587193808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,768,0.05293955405553182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,768,0.03706666827201843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,512,0.02990844514634874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,256,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,512,0.04366933306058248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,128,0.010344888601038191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,256,0.02566577825281355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,256,0.03605599866973029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,64,0.009045333498054082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,128,0.02464444438616435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,2560,0.10645688904656304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,10240,32,0.009316444396972656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,10240,128,0.030152890417310927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,64,0.024711110525661047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,10240,32,0.024432889289326135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,65536,1.059925291273329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,51200,0.7979262140062121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,65536,1.9575093587239583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,16384,0.2668986585405138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,51200,1.5370071199205186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,16384,0.4587893486022949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,65536,1.9032408396402996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,12288,0.20263911618126762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,51200,1.4944044748942058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,16384,0.4938560061984592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,12288,0.3411991066402859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,10240,0.14291377862294516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,12288,0.3748062186770969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,10240,0.2803608841366238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,8192,0.11317155096266006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,10240,0.31148799260457355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,8192,0.220214221212599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,8192,0.2496053377787272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,7168,0.10720711284213597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,6144,0.09092177947362264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,7168,0.1966551144917806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,7168,0.22096178266737196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,5120,0.07381777630911933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,6144,0.16846577326456705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,5120,0.14050666491190592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,4096,0.05903999672995674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,5120,0.16456266244252524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,4096,0.11490044328901504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,3584,0.052188442813025586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,4096,0.13744266827901205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,3584,0.10267288817299737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,6144,0.19206489457024467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,3072,0.0447253319952223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,3584,0.12263111273447673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,2560,0.03965422179963853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,3072,0.09071999788284302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,3072,0.10862488879097833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,2048,0.0322106679280599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,2560,0.07833600044250488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,2560,0.0952862236234877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,1536,0.025801777839660645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,2048,0.06432088878419664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,2048,0.0822497804959615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,1536,0.0513973335425059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,1536,0.06802222463819715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,1024,0.018360889620251127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,768,0.01495822270711263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,1024,0.03868888815244039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,768,0.04869511061244541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,768,0.03306488858328925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,512,0.011529777612951068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,512,0.039028445879618325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,512,0.026383111874262493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,256,0.009392889009581672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,256,0.033724443780051336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,128,0.008007999923494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,256,0.02309155629740821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,128,0.022042666872342426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,1024,0.05412800113360087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,64,0.0074933336840735535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,8192,128,0.030125333203209773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,8192,32,0.007652444144090016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,64,0.02200177808602651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,8192,32,0.022064889470736187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,65536,0.8545360035366483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,51200,0.6516613430447048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,65536,1.7974196539984808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,65536,1.8999608357747395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,51200,1.417501343621148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,16384,0.22965865665011934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,16384,0.4286239941914876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,12288,0.17023021645016143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,16384,0.49323378668891055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,51200,1.4918933444552953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,10240,0.14771289295620388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,12288,0.3165217770470513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,8192,0.11590577496422662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,10240,0.2612355550130208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,12288,0.3754684395260281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,7168,0.10561244355307685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,10240,0.31156444549560547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,8192,0.20474667019314238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,6144,0.0875742236773173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,8192,0.2484382258521186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,7168,0.21946222252315947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,6144,0.15416267183091906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,6144,0.19018756018744576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,5120,0.07394578059514363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,4096,0.05972355604171753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,5120,0.12977155049641928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,5120,0.1620862219068739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,3584,0.05353777938418918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,4096,0.1349946657816569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,4096,0.10675288571251763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,3072,0.045908444457583956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,3584,0.09431555536058213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,3584,0.12111555205451117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,2560,0.039763556586371526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,3072,0.10808889071146648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,3072,0.0849555532137553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,7168,0.178927116923862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,2560,0.07258666886223687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,2048,0.032101333141326904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,2560,0.09455111291673447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,1536,0.025458667013380263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,2048,0.0808195537990994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,2048,0.059454222520192467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,1024,0.019336000084877014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,1536,0.06828177637524076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,768,0.016140444411171805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,1024,0.053284443087048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,1024,0.03634577658441331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,768,0.04673422376314799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,512,0.013763555222087435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,768,0.031236443254682753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,512,0.02473955518669552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,256,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,512,0.038388444317711726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,256,0.021371554997232225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,256,0.033336000310050115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,128,0.009713778065310584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,1536,0.048057778014077075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,64,0.009336888790130615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,7168,128,0.02871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,7168,32,0.009561777942710454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,128,0.021339555581410725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,64,0.020672000116772123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,7168,32,0.02040177749262916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,65536,0.7971359888712565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,51200,0.6515519883897569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,65536,1.6459164089626734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,65536,1.7732435862223308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,16384,0.21175200409359404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,51200,1.295930650499132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,16384,0.4033297697703044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,12288,0.15527378188239202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,16384,0.4614568816290961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,12288,0.29315198792351616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,10240,0.12558310561709934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,12288,0.35039732191297746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,10240,0.2416302098168267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,8192,0.10279999838935004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,10240,0.29094221856859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,51200,1.3902577294243708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,7168,0.09253778060277303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,8192,0.18935733371310762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,8192,0.2329742113749186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,6144,0.07559911410013835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,7168,0.20433689488304985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,7168,0.16456354988945857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,5120,0.06188088655471802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,6144,0.17776089244418672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,6144,0.14294221666124132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,5120,0.12038400438096787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,4096,0.050781332784228854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,5120,0.1518168846766154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,4096,0.09883911079830593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,4096,0.12748532825046116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,3584,0.08753866619533962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,3072,0.03867822223239475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,3584,0.1145697832107544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,3072,0.07571999894248115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,2560,0.03197866678237915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,3072,0.10178933540980022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,2560,0.06722133027182685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,2048,0.02679377794265747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,2560,0.08918222453859116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,3584,0.046374221642812095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,1536,0.021282666259341772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,2048,0.055454221036699086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,1024,0.015923556354310777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,1536,0.06380444765090942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,2048,0.07663910918765597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,768,0.01328000012371275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,1024,0.03394044438997904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,1536,0.04482311010360718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,1024,0.050703111622068614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,512,0.010442666709423065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,768,0.04324711031383938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,256,0.008302222523424361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,768,0.028508444627126057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,512,0.02329066726896498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,128,0.006940444310506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,512,0.03607644306288825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,256,0.02032977839310964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,64,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,256,0.031938665442996554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,128,0.019359999232821994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,6144,32,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,6144,128,0.027615111735132005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,64,0.01927199959754944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,6144,32,0.018974221414989896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,65536,0.6126844618055556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,51200,0.4544382095336914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,65536,1.4977733823988173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,16384,0.1493413315878974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,51200,1.1758826573689778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,65536,1.6429866154988606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,51200,1.291004392835829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,12288,0.11477243900299072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,16384,0.42908888392978245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,10240,0.10079822275373672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,12288,0.2699920071495904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,12288,0.32581867112053764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,8192,0.08287822537951998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,10240,0.2226426601409912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,10240,0.26964444584316677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,7168,0.07428710990481906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,8192,0.1736311117808024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,8192,0.21643555164337158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,16384,0.363009770711263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,7168,0.15208266841040716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,6144,0.06281066603130765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,7168,0.1896151171790229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,5120,0.051069332493676074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,6144,0.1313964393403795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,6144,0.16570044888390437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,5120,0.11115288734436035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,4096,0.04189155499140421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,5120,0.14200533760918513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,4096,0.0907848874727885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,4096,0.11899822288089329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,3584,0.03736622134844462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,3072,0.03279022375742594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,3584,0.08057600259780884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,3584,0.1069271100891961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,2560,0.02786133355564541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,3072,0.06927555799484253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,3072,0.09437155723571777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,2048,0.02308977809217241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,2560,0.05931466817855835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,2048,0.05009599857860141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,1536,0.018440888987647164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,2048,0.07151377863354154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,1536,0.041145778364605375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,1024,0.014184888866212634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,1536,0.06065599785910713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,1024,0.03158044483926561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,768,0.011124444504578909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,1024,0.04836977852715386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,768,0.027148443791601393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,512,0.009346666435400644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,2560,0.08291377623875935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,768,0.04358933369318644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,512,0.02164711058139801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,256,0.0070568886068132185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,512,0.0343537794219123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,128,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,256,0.018952889574898612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,256,0.030890666776233252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,64,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,5120,128,0.026535110341178045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,5120,32,0.006310222049554189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,128,0.018352889352374606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,64,0.018248889181349013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,5120,32,0.01791111131509145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,65536,0.5176595581902398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,51200,0.40155556466844344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,65536,1.3455964194403753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,65536,1.5151146782769098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,51200,1.055246247185601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,16384,0.13858311706119114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,16384,0.33059554629855686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,16384,0.39652355511983234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,12288,0.10228000084559123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,51200,1.1921555201212566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,10240,0.09038311243057251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,12288,0.2460577752855089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,12288,0.3027048905690511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,8192,0.07555022504594591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,10240,0.24946223364935982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,10240,0.20303821563720703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,7168,0.06421599785486858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,8192,0.1589004463619656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,8192,0.19932977358500162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,6144,0.055607113573286265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,7168,0.13905510637495253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,7168,0.17583378156026205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,6144,0.1200595564312405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,5120,0.04697155621316698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,6144,0.1538133356306288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,5120,0.10128266943825616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,4096,0.03808177842034234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,5120,0.13150400585598415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,3584,0.033675554725858904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,4096,0.08312177658081055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,4096,0.11037600040435791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,3072,0.028761777612898085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,3584,0.07320355706744723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,3584,0.0986257791519165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,3072,0.08779822455512153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,2560,0.02474311159716712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,3072,0.06533688969082303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,2048,0.020473778247833252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,2560,0.05505688985188802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,2560,0.07745777898364596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,1536,0.01666933298110962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,2048,0.06636977858013578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,1536,0.037647998995251127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,1536,0.05607822206285265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,1024,0.027814222706688776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,1024,0.042816887299219765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,768,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,768,0.02473955518669552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,768,0.04013599952061971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,512,0.007685333490371704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,512,0.033619556162092425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,256,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,256,0.01753155555990007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,2048,0.04526844289567736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,1024,0.012705778082211813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,256,0.029106666644414265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,128,0.01700533264213138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,4096,128,0.02514666650030348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,512,0.020006222857369315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,64,0.005223999834722943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,32,0.005603555589914322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,64,0.016862221890025668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,4096,32,0.016604445046848722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,4096,128,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,65536,0.4309804439544678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,51200,0.3480737739139133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,65536,1.276112026638455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,51200,0.9996720419989692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,16384,0.11369421746995713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,65536,1.5147440168592665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,16384,0.31485244962904185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,12288,0.08444088697433472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,16384,0.3965911070505778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,51200,1.190599971347385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,10240,0.0748951103952196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,12288,0.3021511236826579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,10240,0.24995377328660753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,10240,0.1937004460228814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,8192,0.05878400140338474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,8192,0.1523182259665595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,8192,0.1997599999109904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,7168,0.04854222138722738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,12288,0.2356631226009793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,6144,0.04181866513358223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,7168,0.17563733789655897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,7168,0.133516444100274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,6144,0.1150968869527181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,5120,0.03650666607750787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,5120,0.09728088643815781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,6144,0.15369421905941433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,4096,0.029322667254341975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,5120,0.13164177205827501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,3584,0.027271999253167048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,4096,0.1094942225350274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,4096,0.08093866374757555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,3072,0.023200889428456623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,3584,0.0719591114256117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,3072,0.06236355834537082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,2560,0.019888000355826486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,3072,0.08827200200822617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,2560,0.053010665708118014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,2560,0.07726577917734782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,2048,0.016693333784739178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,2048,0.0436462230152554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,2048,0.06627555688222249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,1536,0.013855111267831592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,3584,0.09810044368108113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,1024,0.010668444136778513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,1536,0.03614577651023865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,1536,0.056877334912618004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,1024,0.027101332942644756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,768,0.00870311094654931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,512,0.007672888537247975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,1024,0.042928887738121875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,768,0.023411555422676936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,256,0.006270222365856171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,768,0.038908445172839694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,512,0.01963200006220076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,128,0.005633777628342311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,512,0.033287998702791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,256,0.016843555702103507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,256,0.028797331783506606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,64,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3584,32,0.005692444327804778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3584,128,0.02517333295610216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,64,0.016159999701711867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,128,0.016536000702116225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3584,32,0.016192888220151264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,51200,0.299507564968533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,65536,0.3752088811662462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,65536,1.19541507297092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,51200,0.9354391098022461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,16384,0.09998133447435166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,16384,0.298379553688897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,65536,1.5151661766899958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,51200,1.18850220574273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,16384,0.3962577713860406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,12288,0.0776817798614502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,10240,0.06730222039752536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,10240,0.18352977434794107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,12288,0.22343378596835664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,8192,0.05040355523427328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,12288,0.3014764520857069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,7168,0.0456080006228553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,8192,0.19896799988216826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,8192,0.14425422085656062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,6144,0.03818488783306546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,7168,0.12558666865030924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,7168,0.17469955815209282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,5120,0.03267022305064731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,6144,0.10905866490470038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,6144,0.15283555454678005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,4096,0.028427554501427546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,5120,0.13164356019761828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,10240,0.24894221623738608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,3584,0.025066667132907446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,4096,0.07566577858395047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,4096,0.1093920005692376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,3584,0.06750310791863336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,3072,0.022403554783927068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,3584,0.0986951125992669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,3072,0.05841866466734144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,5120,0.0920320020781623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,3072,0.08791110912958781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,2560,0.018815999229749043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,2048,0.017080000705189176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,2560,0.04935466580920749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,2560,0.07730666796366374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,1536,0.01329422245422999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,2048,0.04071111149258084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,2048,0.06572000185648601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,1024,0.009338666995366415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,1536,0.034266667233573064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,1024,0.02607911162906223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,768,0.008343111309740279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,1024,0.04264266623391045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,768,0.02239466706911723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,512,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,768,0.038791110118230186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,512,0.018749333090252347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,256,0.005920889063013925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,512,0.03233244352870517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,256,0.016584000653690763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,128,0.004932444542646408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,1536,0.054719110329945884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,256,0.02861777941385905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,128,0.01585155559910668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,64,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,3072,32,0.00499555551343494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,3072,128,0.024540444215138752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,64,0.01552088889810774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,3072,32,0.015593777100245157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,65536,0.29785333739386666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,51200,0.24110844400193956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,65536,1.1331769095526802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,51200,0.887454244825575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,16384,0.08100889126459758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,65536,1.3877858055962458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,16384,0.28461511929829914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,16384,0.3647777769300673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,51200,1.0882800420125325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,12288,0.21375021669599745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,12288,0.27753867043389213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,10240,0.049900445673200816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,10240,0.1759199963675605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,8192,0.04027022255791558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,10240,0.22991644011603463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,12288,0.060180445512135826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,8192,0.13830488257937962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,7168,0.037631111012564764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,8192,0.18332800600263807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,6144,0.03240977724393209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,7168,0.12040089236365424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,7168,0.16116355525122747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,5120,0.02777422302299076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,6144,0.1039724416202969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,6144,0.14028178320990667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,4096,0.02269866731431749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,5120,0.08834489186604817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,5120,0.12053333388434516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,4096,0.0725839998986986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,3584,0.02106222179200914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,3072,0.018383110562960308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,3584,0.06432711415820651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,4096,0.10079022248586018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,3584,0.09130666653315227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,2560,0.016490666402710807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,3072,0.05534933341874016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,2048,0.013752889302041797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,3072,0.08091644446055095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,2560,0.048285331990983754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,2048,0.03875644339455499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,2048,0.05946844153934055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,1536,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,1024,0.008651555412345463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,1536,0.031735999716652766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,1024,0.02477866742346022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,1024,0.03973866502443949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,768,0.0074240002367231585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,768,0.02134399943881565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,512,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,768,0.03675377700063918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,512,0.01787644459141625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,256,0.0052728888889153796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,512,0.03092977735731337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,2560,0.07150044706132677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,256,0.0155102229780621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,128,0.0046453335218959385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,256,0.02781244450145298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,64,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,1536,0.04997244477272034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,128,0.0148044443792767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2560,32,0.004605333424276776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2560,128,0.024136000209384497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,64,0.014909333652920194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2560,32,0.014474666780895658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,65536,0.2786640061272515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,51200,0.2220462163289388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,65536,1.0481066173977323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,51200,0.8256470892164441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,16384,0.07317777474721272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,65536,1.3862151039971247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,16384,0.26698843638102215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,12288,0.05467111203405592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,16384,0.3645182185702854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,51200,1.0887502034505208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,12288,0.19949600431654188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,10240,0.047167999876870044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,10240,0.16471289263831243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,8192,0.03769866625467936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,10240,0.22930754555596244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,8192,0.12921867105695936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,7168,0.03344355689154731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,8192,0.1831359995736016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,7168,0.11220088269975449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,6144,0.02944177720281813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,7168,0.1608417828877767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,6144,0.09715822007921006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,12288,0.27724356121487087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,5120,0.02551288902759552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,6144,0.1405457788043552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,4096,0.022080888350804646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,5120,0.12022222412957086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,5120,0.08249866962432861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,3584,0.01941511034965515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,4096,0.10052533282174005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,4096,0.0672631131278144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,3072,0.017824000782436795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,3584,0.059472892019483775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,3584,0.090302222304874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,2560,0.01519555515713162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,3072,0.051121777958340116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,3072,0.08135288953781128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,2560,0.04331733451949226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,2048,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,2560,0.07131910986370511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,2048,0.03637866510285272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,1536,0.009576000273227692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,2048,0.06134044461780124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,1024,0.00739911115831799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,1536,0.02996088729964362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,1536,0.04999555481804741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,768,0.00665866666369968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,1024,0.0394755568769243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,1024,0.023411555422676936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,512,0.00574933323595259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,768,0.03622311022546556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,768,0.020430222153663635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,256,0.004990222139490975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,512,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,512,0.03049955434269375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,128,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,256,0.014901333385043673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,256,0.027544000082545813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,128,0.014581332604090372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,64,0.003976888954639435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,2048,32,0.004335111214054955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,64,0.014352889524565803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,65536,0.19891556104024252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,2048,32,0.014473777678277759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,2048,128,0.023167999254332647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,65536,0.9878738191392687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,51200,0.15739288594987658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,16384,0.05386399891641405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,65536,1.3867929246690538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,16384,0.25432801246643066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,51200,0.7764968872070312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,12288,0.03914133376545376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,16384,0.36449334356519913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,12288,0.19068888823191324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,10240,0.034316443734698825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,51200,1.087731573316786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,8192,0.02892000145382351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,12288,0.2773173385196262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,10240,0.15727821985880533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,7168,0.02388977838887109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,10240,0.22951555252075195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,8192,0.12242755624983047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,6144,0.02113866640461816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,8192,0.18335555659400093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,7168,0.10708177751964992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,5120,0.01833599971400367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,7168,0.16112177901797825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,6144,0.09197511275609334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,4096,0.01569600072171953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,6144,0.1401537789238824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,5120,0.07809155517154269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,3584,0.014360000689824423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,4096,0.06349955664740668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,5120,0.12041866779327393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,4096,0.10018489095899795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,3072,0.013047999805874295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,3584,0.055758224593268506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,3584,0.09014399846394856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,2560,0.011376000112957425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,2560,0.041137778096728854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,2560,0.07021155622270372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,3072,0.04902844296561348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,3072,0.08135555850134955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,2048,0.03406577640109592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,2048,0.060738669501410596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,1536,0.008729777402347988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,1536,0.049272000789642334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,1536,0.028562668297025893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,1024,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,1024,0.022389334109094407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,1024,0.03936800029542711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,768,0.005975111077229182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,2048,0.009671111073758867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,512,0.0052435555391841466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,768,0.01925155520439148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,512,0.016233777006467182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,768,0.03600622216860453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,256,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,512,0.030215998490651447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,128,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,256,0.014475555883513557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,128,0.013560000393125745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,64,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,128,0.024400000770886738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1536,32,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,64,0.013237333132161034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1536,32,0.013935999737845527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,65536,0.14269155926174587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1536,256,0.027494221925735474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,51200,0.10677866803275214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,65536,0.9056853188408746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,51200,0.7113315794203016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,16384,0.03738133443726434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,65536,1.2545892927381728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,16384,0.23575822512308756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,12288,0.02958044409751892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,16384,0.33075200186835396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,51200,0.9866569307115344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,12288,0.1772986650466919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,10240,0.024432889289326135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,12288,0.2523750993940565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,10240,0.1452853282292684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,8192,0.021020443903075323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,10240,0.2087928851445516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,8192,0.11313066217634414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,7168,0.018160000443458557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,8192,0.16632710562811956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,6144,0.01662133302953508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,7168,0.0984133349524604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,7168,0.1459208859337701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,6144,0.08515289094712998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,5120,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,5120,0.07148000266816881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,5120,0.10954044262568156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,4096,0.011220444407727985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,4096,0.09104444583257039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,4096,0.058240003055996366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,3584,0.010418666733635796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,6144,0.12667377789815268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,3584,0.08139999707539876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,3584,0.05152533451716105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,3072,0.009445333646403419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,2560,0.007995555798212687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,3072,0.04467022087838915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,2560,0.03826489051183065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,3072,0.07413866784837511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,2048,0.006963555183675554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,2048,0.05379377802213033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,2560,0.0641182197464837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,1536,0.0062871111763848206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,2048,0.03161066770553589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,1024,0.0053013331360287136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,1536,0.04554222358597649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,1536,0.026291555828518335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,768,0.004927111168702443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,1024,0.03628266519970364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,512,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,768,0.03327733278274536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,1024,0.020627554919984605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,768,0.018214222457673814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,512,0.015472888946533203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,512,0.028512001037597656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,128,0.0035911111368073356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,256,0.025721778472264607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,256,0.013451555536852943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,64,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,128,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,32,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,1024,128,0.02199999988079071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,65536,0.12159200509389241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,32,0.013045333325862885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,1024,64,0.012616000241703458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,51200,0.09731200006273057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,1024,256,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,65536,0.8969848420884875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,16384,0.034918222162458636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,51200,0.7037422392103406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,65536,1.252943992614746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,16384,0.2334524525536431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,12288,0.024160888459947374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,16384,0.33061332172817653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,51200,0.9862231148613824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,12288,0.17538489235772026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,10240,0.020240889655219186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,12288,0.25199733840094674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,10240,0.14459466934204102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,8192,0.01759466694460975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,10240,0.20857955349816215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,7168,0.015574221809705099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,8192,0.11197866333855523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,8192,0.16558310720655653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,6144,0.014225777652528552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,7168,0.14533955521053737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,7168,0.09761066569222344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,5120,0.012163555456532372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,6144,0.08404622475306193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,6144,0.12650044759114584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,4096,0.01018311083316803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,5120,0.10891999800999959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,5120,0.07107200225194295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,3584,0.009467555416954888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,4096,0.0914435585339864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,3584,0.05023733443684048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,3072,0.008709333009190029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,3584,0.08146755562888251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,3072,0.044143110513687134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,3072,0.07329155339135064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,2560,0.007747555772463481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,4096,0.05719377597173055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,2048,0.0069546666410234236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,2560,0.03689600030581156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,2560,0.06485155555937025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,1536,0.0063742221229606206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,2048,0.031049778064092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,2048,0.05349777804480659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,1024,0.0053688887920644544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,1536,0.04520533482233683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,1536,0.02644177774588267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,768,0.0053102220926019884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,1024,0.03605511122279697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,1024,0.020415110720528495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,512,0.004360888981156879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,768,0.03291911217901442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,768,0.01830222209294637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,256,0.003942222230964237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,512,0.02803200152185228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,512,0.014907555447684394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,128,0.003630222131808599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,256,0.02509244448608822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,256,0.013447999954223633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,64,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,768,128,0.022364444202846948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,128,0.012878222597969903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,768,32,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,65536,0.08684355682796902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,64,0.012712000144852532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,768,32,0.012167111039161682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,51200,0.07051377826266818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,65536,0.8580408626132541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,51200,0.674015998840332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,16384,0.025937777426507738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,65536,1.2435102462768555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,16384,0.22421156035529244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,12288,0.019170666734377544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,51200,0.9761386447482638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,16384,0.3278488847944471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,10240,0.01608088943693373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,12288,0.16894043816460502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,12288,0.24915377298990884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,8192,0.014113777213626437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,10240,0.13902310530344644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,10240,0.2056257724761963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,7168,0.012649777862760754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,8192,0.10753777954313491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,8192,0.16438133186764187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,7168,0.09359555774264866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,7168,0.1437493297788832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,6144,0.011123555401961008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,5120,0.009717333647939894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,6144,0.12582488854726157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,6144,0.080677330493927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,4096,0.008415111237102086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,5120,0.06816711028416951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,5120,0.10809244712193807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,3584,0.007998222278224098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,4096,0.09018044339285956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,4096,0.05458844370312161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,3584,0.0470053321785397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,3072,0.007375111182530721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,3072,0.041690667470296226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,2560,0.006688000013430913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,3584,0.08136800262663099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,3072,0.07241155703862508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,2048,0.005985777825117111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,2560,0.03492710987726847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,2560,0.06376266479492188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,1536,0.005294222384691238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,2048,0.029688000679016113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,1536,0.025076443950335186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,2048,0.05350133445527819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,1536,0.04478311207559374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,1024,0.004552888787455029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,768,0.0041911109454101985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,1024,0.01980444457795885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,1024,0.03569777806599935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,512,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,768,0.017656889226701524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,768,0.03324000040690104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,256,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,512,0.014771555860837301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,512,0.027898665931489732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,256,0.012732444538010491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,128,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,128,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,128,0.021339555581410725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,64,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,32,0.00328177772462368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,512,32,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,65536,0.05841244591606987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,512,256,0.024713777833514746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,512,64,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,51200,0.04703911145528158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,65536,0.8192755381266276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,51200,0.6425191031561958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,16384,0.017755554782019723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,65536,1.2412755754258897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,16384,0.21404355102115205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,12288,0.014775110615624322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,16384,0.3272675673166911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,51200,0.9762658013237847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,10240,0.014271999398867289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,12288,0.16125778357187906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,12288,0.24890576468573677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,8192,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,10240,0.13219288984934488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,10240,0.2050328916973538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,7168,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,8192,0.10245511266920303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,8192,0.16397510634528265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,6144,0.009735999835862054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,7168,0.08875288565953572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,7168,0.1437564425998264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,5120,0.008618666893906063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,6144,0.1253368854522705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,6144,0.07660800218582153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,4096,0.007617777420414819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,5120,0.06418666574690077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,5120,0.10792266660266453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,3584,0.006912888752089606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,4096,0.05153510967890421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,4096,0.08952977922227646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,3072,0.006643555644485686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,3584,0.045143110884560474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,3584,0.08039377795325385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,2560,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,3072,0.037997331884172224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,3072,0.06979733043246798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,2048,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,2560,0.060912887255350746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,2560,0.03386044502258301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,1536,0.004918222212129169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,2048,0.02847644355561998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,2048,0.05272444420390659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,1024,0.004173333446184794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,1536,0.044532444741990834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,1536,0.02404088940885332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,768,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,1024,0.018903111418088276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,1024,0.035586665074030556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,512,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,768,0.03286844491958618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,768,0.016861332787407767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,256,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,512,0.014112000664075216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,512,0.027464000715149775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,128,0.003167111012670729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,256,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,256,0.024698666400379602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,64,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,128,0.011704000333944956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,256,32,0.00315022220214208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,256,128,0.020787555310461257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,65536,0.044333332114749484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,64,0.011698666546079846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,256,32,0.011647111011876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,51200,0.03705777724583944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,65536,0.8185351159837512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,51200,0.6424995528327094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,16384,0.011572444604502784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,16384,0.2137671046786838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,12288,0.010001777774757808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,16384,0.32678045166863334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,12288,0.1611893309487237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,10240,0.009288888838556077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,12288,0.24919022454155815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,10240,0.13236267036861843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,8192,0.010065777434243096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,10240,0.205366227361891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,51200,0.9769715203179253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,7168,0.011221333510345884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,8192,0.10206044382519192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,8192,0.1640053325229221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,65536,1.2413413789537218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,6144,0.009731555150614845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,7168,0.08843021922641331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,7168,0.14360622564951578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,5120,0.008639111287064022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,6144,0.07610488600201078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,6144,0.12481510639190674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,4096,0.007636444436179266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,5120,0.06335288948482938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,5120,0.10714577966266209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,3584,0.006918222126033571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,4096,0.08927288982603286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,3584,0.04423377911249796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,3072,0.0064479998416370815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,3584,0.08034666379292806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,3072,0.038902223110198975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,3072,0.07174044185214572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,2560,0.0336408880021837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,4096,0.05105244451098972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,2560,0.0629591080877516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,2048,0.00526577772365676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,2048,0.05272000034650167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,1536,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,1536,0.024069334069887798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,2560,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,1024,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,1536,0.04412710997793409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,1024,0.018958222534921434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,768,0.003650666524966558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,1024,0.03535999854405721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,768,0.016948444975747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,768,0.03260533346070184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,2048,0.028200000524520874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,512,0.014175110393100314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,256,0.0034186666210492453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,512,0.027499554885758296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,256,0.012142221960756512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,128,0.00295733329322603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,256,0.024098666177855596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,128,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1792,128,128,0.020671111014154222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,32,0.0033004445334275565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,64,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,128,512,0.003495111233658261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,65536,0.04279999931653341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,128,32,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,51200,0.03567022085189819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,16384,0.009745777481132084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,16384,0.2137662172317505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,12288,0.008954666554927826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,51200,0.6423119968838161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,65536,0.8179715474446615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,10240,0.010008888940016428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,8192,0.010062221851613786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,12288,0.1610648896959093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,7168,0.009009777671760982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,10240,0.13174133830600315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,8192,0.10168977578481038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,6144,0.00870222184393141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,7168,0.08828177716996934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,5120,0.008000888758235509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,4096,0.007275555696752336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,6144,0.07635377513037787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,5120,0.06398844718933105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,3584,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,4096,0.05097511079576281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,3072,0.0063866666621632045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,2560,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,3584,0.04455822375085619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,2048,0.005224888937340842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,3072,0.03802666730350918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,1536,0.004578666554556953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,2560,0.032567110326555036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,1024,0.0042053332759274375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,2048,0.028200000524520874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,768,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,1536,0.02401688860522376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,512,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,1024,0.018573333819707234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,256,0.0031439999325407874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,512,0.014239999983045789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,128,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,256,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,64,0.0028106667515304354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,128,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,64,32,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,64,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,65536,0.04006577862633599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,32,0.011375111010339526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,51200,0.03326311045222812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,64,768,0.016879111528396606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,16384,0.009324444664849175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,51200,0.6419315338134766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,65536,0.8185831175910102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,12288,0.008309333688682979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,10240,0.007617777420414819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,16384,0.21354399787055123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,12288,0.16087022092607287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,10240,0.1320026715596517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,8192,0.0073315559162033936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,7168,0.010767111347781287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,8192,0.10187200042936538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,6144,0.010011555420027839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,7168,0.08816355466842651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,5120,0.009110222260157267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,6144,0.07598666350046794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,4096,0.00831200016869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,5120,0.06328088707394071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,3584,0.006993778049945831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,4096,0.05072888731956482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,3072,0.006251555350091722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,3584,0.04420444369316101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,2560,0.005938666562239329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,3072,0.03858310977617899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,2048,0.005575110928879843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,2560,0.0325617790222168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,1536,0.004597333156400257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,2048,0.02829333477550083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,1024,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,1536,0.02369955513212416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,768,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,1024,0.018914666440751817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,512,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,768,0.0166320006052653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,512,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,128,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,256,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,64,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,128,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1792,32,32,0.002573333266708586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,64,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1792,32,32,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,16384,1.5582168367173936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,51200,4.644294314914279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,16384,2.242097854614258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,12288,1.1411795086330838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,10240,0.9251777860853406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,12288,1.578585730658637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,10240,1.2858835856119792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,51200,7.587390475802952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,8192,0.7530693478054471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,7168,0.6723146438598633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,8192,1.0080177519056532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,6144,0.6033644676208496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,7168,0.8734453519185384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,5120,0.47576268513997394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,6144,0.7802506552802192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,5120,0.6185279952155219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,3584,0.35257776578267414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,4096,0.49672709570990664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,3072,0.30301332473754883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,3584,0.43556443850199383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,2560,0.2578950987921821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,4096,0.3847093317243788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,2560,0.3190560075971815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,3072,0.3754453394148085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,2048,0.24118489689297148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,1536,0.1901679966184828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,1024,0.13500711652967665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,2048,0.26188000043233234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,768,0.10680355628331502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,1536,0.20752978324890137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,512,0.0802417794863383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,1024,0.15821777449713811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,256,0.0540595551331838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,768,0.12639821900261775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,128,0.04088000125355191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,64,0.03487111131350199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,512,0.10927289062076145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,256,0.09782310989167954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,65536,32,0.036657777097490095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,128,0.09326489104164971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,64,0.09299111366271973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,65536,32,0.09366044733259414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,51200,3.7734523349338107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,65536,4.902033064100477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,16384,1.2162595325046115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,12288,0.9232301712036133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,51200,5.987005445692275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,16384,1.7737048467000325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,10240,0.847751087612576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,12288,1.2459066179063585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,65536,7.730007595486111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,8192,0.7018017768859863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,10240,1.0229920281304252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,7168,0.6277919875250922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,8192,0.8078675799899631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,7168,0.7005786895751953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,6144,0.5266088909573025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,5120,0.4236995644039578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,6144,0.5968355602688259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,4096,0.351947546005249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,3584,0.2865377797020806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,5120,0.49524709913465714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,4096,0.4172364340888129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,3584,0.37457866138882107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,3072,0.23440800772772896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,2560,0.197906666331821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,3072,0.3008382320404053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,2048,0.16113600465986463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,2560,0.256751987669203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,1536,0.12462045086754693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,2048,0.21167643864949545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,1024,0.09078666898939346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,768,0.07368266582489014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,1536,0.16734577549828422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,512,0.052728000614378184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,1024,0.12820977634853786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,256,0.04345155424541897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,768,0.10207555691401164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,128,0.03353066576851739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,512,0.08891199694739448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,256,0.07955733272764418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,128,0.0753600001335144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,32,0.03459022111362881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,64,0.0757804446750217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,51200,32,0.0759155551592509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,51200,64,0.03263466556866964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,65536,1.4645324283176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,51200,1.1510000228881836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,16384,0.3656986554463704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,65536,2.838494194878472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,16384,0.6466079817877876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,12288,0.29426489935980904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,10240,0.23643377092149523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,51200,2.190442615085178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,12288,0.4684346516927083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,8192,0.1906648874282837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,7168,0.17003022299872506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,10240,0.3855653338962131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,8192,0.3054684533013238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,6144,0.1579795546001858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,5120,0.12422755029466416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,7168,0.27090488539801705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,6144,0.2328746583726671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,4096,0.10190311405393813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,3584,0.09138044383790757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,5120,0.19472266568077934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,3072,0.08641778098212348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,3584,0.14040444956885442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,2560,0.06769866413540311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,3072,0.12233511606852214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,2048,0.058239108986324735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,2560,0.10582933160993789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,2048,0.0883057779735989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,1536,0.04486666785346138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,1024,0.032666666640175715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,1536,0.07007733318540785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,768,0.02626755502488878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,1024,0.05389066537221273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,512,0.01957066688272688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,4096,0.15839021735721165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,256,0.015796444482273526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,512,0.036405334870020546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,128,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,256,0.03230399886767069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,64,0.011030221978823343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,128,0.03019022279315525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,16384,32,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,64,0.03019555409749349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,32,0.02985066506597731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,16384,768,0.04318399892912971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,65536,1.1169875462849934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,51200,0.8968462414211698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,16384,0.29817244741651744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,12288,0.22981156243218315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,65536,2.279272927178277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,16384,0.5227448675367568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,10240,0.17849600315093994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,8192,0.14907999833424887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,51200,1.7546017964680989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,10240,0.3150248792436388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,7168,0.13269510534074572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,8192,0.2485128773583306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,6144,0.10872533586290146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,5120,0.09280266364415486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,7168,0.21966132852766249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,6144,0.18934310807122123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,4096,0.08580888642205133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,12288,0.4133271111382379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,5120,0.159990217950609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,4096,0.13044977188110352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,3072,0.0662355555428399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,2560,0.05443466703097025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,3584,0.11652355723910862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,3072,0.10142755508422852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,2048,0.0449848903550042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,1536,0.03537688983811273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,2560,0.0884737769762675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,2048,0.07292089197370741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,1024,0.025440000825458105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,768,0.020407110452651978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,1536,0.05752444267272949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,512,0.016024889217482675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,1024,0.04492977923817105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,768,0.03648533423741659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,256,0.0147386673423979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,512,0.030194666650560167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,128,0.01000088867213991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,256,0.02641955514748891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,64,0.00902844468752543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,128,0.025412445267041523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,32,0.0090524446633127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,12288,3584,0.07593244314193726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,64,0.025417778227064345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,12288,32,0.025055999557177227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,51200,0.7458560201856824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,65536,0.9748640060424805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,16384,0.25861777199639213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,16384,0.463031980726454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,65536,1.993130577935113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,12288,0.18685511747996011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,10240,0.1628471083111233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,51200,1.5458186467488606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,12288,0.33984356456332737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,8192,0.12951733006371394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,7168,0.1144755548901028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,8192,0.21997688876258004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,10240,0.2801644537183973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,6144,0.0971520013279385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,5120,0.08291466368569268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,7168,0.1939342286851671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,4096,0.07037066751056247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,6144,0.1687671078575982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,5120,0.14131822850969103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,3584,0.06074755721622043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,4096,0.11638666523827447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,3072,0.05332977904213799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,3584,0.10391910870869954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,2560,0.045831110742357045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,2048,0.03829688827196757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,3072,0.09154666794670953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,1536,0.030996445152494643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,2560,0.0788408915201823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,1024,0.023600889576805964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,2048,0.06647644440333049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,1536,0.052054223087098866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,768,0.020238222347365487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,1024,0.04031022058592902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,512,0.015220445063379077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,256,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,768,0.032990223831600614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,128,0.009670221971140968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,512,0.027172444595230952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,64,0.008761777646011775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,128,0.023024888502226934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,10240,32,0.008655110994974772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,64,0.02313511073589325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,32,0.022760889596409265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,65536,0.9541698031955295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,10240,256,0.02436711059676276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,51200,0.7472489145067004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,65536,1.7185911602444117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,16384,0.24584621853298613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,12288,0.18844000498453775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,16384,0.40640711784362793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,51200,1.3220710754394531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,10240,0.1373937792248196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,8192,0.1087484425968594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,10240,0.2436906761593289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,7168,0.10286844438976711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,8192,0.19240710470411512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,6144,0.09068444702360366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,7168,0.1702693303426107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,5120,0.0870364440812005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,6144,0.14679733912150064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,4096,0.058451553185780845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,3584,0.056348443031311035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,12288,0.2971511152055529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,5120,0.1240053309334649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,3072,0.04472088813781738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,4096,0.1014702253871494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,2560,0.040439999765819974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,2048,0.03146044413248698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,2560,0.06943644417656793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,3072,0.08070400026109484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,1536,0.027928001350826685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,2048,0.056836446126302086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,1024,0.01794311073091295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,768,0.015303111738628812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,1024,0.03580177823702494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,1536,0.04629244407018026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,768,0.029169778029123943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,256,0.009089777866999308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,512,0.024933333198229473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,256,0.021624000536070928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,128,0.008408000071843466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,64,0.007801777786678738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,128,0.021019554800457422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,3584,0.0914622214105394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,32,0.007611555357774098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,32,0.020303999384244282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,8192,64,0.021253334151373968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,8192,512,0.012387555506494312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,65536,0.7262133492363824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,51200,0.5615208943684896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,16384,0.19111288918389213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,16384,0.3791680071089003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,12288,0.13956177234649658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,65536,1.5920755598280165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,51200,1.2291777928670247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,10240,0.12432800398932563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,12288,0.2762266794840495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,8192,0.09193066755930583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,7168,0.0810826685693529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,10240,0.2255715529123942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,6144,0.07067910830179851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,8192,0.17866044574313691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,5120,0.060658666822645396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,7168,0.15753866566552058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,4096,0.050958222813076444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,6144,0.13632799519432917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,5120,0.11558311515384251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,3584,0.04452711012628344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,3072,0.03898755709330241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,4096,0.09473333093855117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,2560,0.03340355555216471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,3584,0.08427021900812785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,2048,0.027048889133665297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,3072,0.07515288723839654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,1536,0.022270222504933674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,2560,0.06391288836797078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,1024,0.01627200014061398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,2048,0.053363555007510714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,768,0.01329422245422999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,1024,0.033302221033308245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,512,0.010455110834704505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,768,0.027645332945717707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,256,0.00831911133395301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,512,0.023023999399609033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,128,0.007354666789372762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,256,0.020651555723614164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,64,0.0069653333889113525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,128,0.01965688831276364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,7168,32,0.006949333681000604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,64,0.019726221760114033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,32,0.019323555959595572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,7168,1536,0.042340444193945996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,65536,0.613426685333252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,51200,0.44585688908894855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,16384,0.14834754996829563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,16384,0.3509608904520671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,65536,1.437710232204861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,12288,0.11287911732991536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,51200,1.1190453635321724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,10240,0.09545066621568467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,8192,0.07792177465226915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,12288,0.25485777854919434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,7168,0.07224444548288982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,10240,0.20819377899169922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,8192,0.1645644505818685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,6144,0.061536000834570996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,5120,0.052730666266547314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,6144,0.12607378429836696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,5120,0.10707200235790676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,4096,0.04276177618238661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,3584,0.03840533230039809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,4096,0.08774755398432414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,3584,0.07707022296057807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,3072,0.034007999632093645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,2560,0.029481778542200725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,3072,0.0664222240447998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,2048,0.024657777614063684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,7168,0.14482043849097356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,1536,0.02091288897726271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,2048,0.04857066604826185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,1024,0.01609155535697937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,1536,0.03881866733233134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,768,0.0135057775510682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,1024,0.031112889448801678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,512,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,768,0.025776000486479864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,256,0.008349333372380998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,512,0.021384000778198242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,128,0.007350222104125553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,256,0.019262222780121695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,2560,0.05911822451485527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,64,0.0063279999627007385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,128,0.018610666195551556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,6144,32,0.00692622239391009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,32,0.018230222993426852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,65536,0.48019732369316953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,51200,0.3745288848876953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,6144,64,0.01854311095343696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,65536,1.3151813083224826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,16384,0.3168124357859294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,12288,0.09477421972486709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,51200,1.0195093154907227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,10240,0.09432622459199692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,12288,0.23433245552910698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,8192,0.07333688603507148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,10240,0.19123466809590658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,16384,0.12916710641649035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,7168,0.06674489047792223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,6144,0.058435552650027804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,8192,0.1513351069556342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,7168,0.133952882554796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,5120,0.0467217763264974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,6144,0.11555643876393636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,4096,0.03808710972468058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,5120,0.09898489051394993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,4096,0.08021066586176555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,3584,0.07097688648435804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,3072,0.030204445123672485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,3072,0.06405866808361478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,2560,0.025723555021815833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,2560,0.0532168878449334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,2048,0.021613332960340712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,3584,0.033597333563698664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,1536,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,2048,0.04568799999025133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,1024,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,1536,0.036954667833116316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,768,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,1024,0.028365333875020344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,512,0.008722667064931657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,768,0.0240631103515625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,256,0.006796444455782573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,512,0.02037333283159468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,128,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,256,0.01808177762561374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,64,0.00566133318675889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,128,0.017284444636768766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,5120,32,0.005956444475385878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,64,0.01756533318095737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,5120,32,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,65536,0.4576097594367133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,51200,0.3658577866024441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,65536,1.1744026607937283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,16384,0.11816444661882188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,51200,0.9127697414822049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,16384,0.29004621505737305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,12288,0.09054844246970283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,10240,0.07759288946787517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,12288,0.21427288320329455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,8192,0.06269955635070801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,10240,0.17527466350131563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,7168,0.05605510870615641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,8192,0.13845245043436685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,6144,0.055454221036699086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,7168,0.12299466133117676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,5120,0.04594399862819248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,6144,0.10603733195198907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,4096,0.03830044468243917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,4096,0.07456533114115398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,3584,0.033129778173234724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,3072,0.030244443151685927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,3584,0.06594044632381864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,2560,0.026719111535284255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,3072,0.05762044588724772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,2560,0.04844355583190918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,2048,0.02051199972629547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,1536,0.0163102216190762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,2048,0.04096622268358866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,5120,0.09069510963228013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,1536,0.03401244348949856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,1024,0.02574666672282749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,768,0.010149333212110732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,512,0.008083555433485243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,768,0.022241777843899194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,256,0.006256888724035687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,512,0.019352000620630052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,128,0.005680000202523337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,128,0.016568889220555622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,1024,0.012759110993809171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,64,0.005119110975000593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,4096,32,0.0053084443012873335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,64,0.01586844523747762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,32,0.0162435554795795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,65536,0.3446462154388428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,4096,256,0.016858667135238647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,51200,0.27934667799207896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,16384,0.09320088889863755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,12288,0.07105955812666151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,16384,0.2755511071946886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,10240,0.06130044327841865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,12288,0.20377066400316027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,51200,0.870006243387858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,8192,0.050071110328038536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,65536,1.1107688479953342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,7168,0.043232887983322144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,6144,0.037956443097856306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,10240,0.16752978165944418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,8192,0.13257243898179796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,5120,0.03291555576854282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,7168,0.11757244004143609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,6144,0.10216533475452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,3584,0.024000888069470722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,5120,0.08639733658896552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,3072,0.022460444105996027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,4096,0.07170577843983968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,3584,0.0632622241973877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,2560,0.018561777141359117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,3072,0.05454666746987236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,2048,0.01612355477280087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,1536,0.013221333424250284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,2560,0.04664088951216804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,2048,0.03915289044380188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,1024,0.009597332941161262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,1536,0.03244355652067397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,768,0.00796888851457172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,1024,0.02478755513827006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,512,0.006578666468461354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,768,0.021574222379260596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,256,0.005945777727497949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,4096,0.02717155549261305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,128,0.005227555417352253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,512,0.018232888645595975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,64,0.00462400002612008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,256,0.016507556041081745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,128,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,64,0.015526221858130561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3584,32,0.01536799967288971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,65536,0.29270754920111763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,51200,0.22526221805148652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3584,32,0.004882666799757215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,65536,1.0356942282782662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,16384,0.07958666483561198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,12288,0.06012088722652859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,51200,0.8166009055243598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,10240,0.05398133397102356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,16384,0.2607768906487359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,12288,0.19188533888922799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,8192,0.041639112763934664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,7168,0.03735111157099406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,10240,0.15762133068508571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,6144,0.03250844279925028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,8192,0.12453244792090522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,5120,0.027489778068330552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,7168,0.10962400171491836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,4096,0.02346044447686937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,6144,0.09568888611263698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,3584,0.02127911150455475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,5120,0.08131200075149536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,3072,0.018605333235528734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,4096,0.06620533598793878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,2560,0.01680088871055179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,3584,0.05878577629725138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,2048,0.014278221461508008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,3072,0.05121333400408427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,1536,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,2560,0.043506665362252124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,2048,0.03648799988958571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,1536,0.02983022067281935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,768,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,1024,0.02367377777894338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,512,0.0069208890199661255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,768,0.020344000723626878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,256,0.005548444473081165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,512,0.017511111166742113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,128,0.004875555634498596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,256,0.015863110621770222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,64,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,128,0.015164444843928019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,32,0.004920000003443824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,64,0.015150222513410779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,3072,32,0.01479111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,3072,1024,0.009016888837019602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,65536,0.27098133828904897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,51200,0.20416000154283312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,16384,0.06990755266613431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,16384,0.24854487842983672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,12288,0.05345244540108574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,51200,0.7692444589402941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,65536,0.9770541720920138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,10240,0.04574044545491537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,8192,0.03699644406636556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,12288,0.18410844273037383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,10240,0.15005244149102107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,7168,0.03296622302797105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,8192,0.11886578136020237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,6144,0.028119110398822363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,5120,0.02420888841152191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,7168,0.10420711172951592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,4096,0.02052888936466641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,6144,0.0913155542479621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,3584,0.018244443668259513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,5120,0.07739733325110541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,3072,0.016473778420024447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,4096,0.06297689014010958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,2560,0.014469333820872836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,3584,0.055642664432525635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,2048,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,3072,0.04896266592873467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,1536,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,2560,0.041659557157092623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,1024,0.008405333591832055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,2048,0.03457866774664985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,1536,0.028347555134031508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,1024,0.022650667362742957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,512,0.0059075554211934405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,768,0.019409777389632333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,256,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,512,0.016827555166350473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,128,0.00452533322903845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,256,0.015185778339703878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,128,0.01443822185198466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,64,0.0041866666740841335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,64,0.014111111561457316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,32,0.0041884444653987885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2560,768,0.0069679998689227635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2560,32,0.014129777749379476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,65536,0.25020623207092285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,51200,0.19672621621025932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,65536,0.9268364376491971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,16384,0.0657600031958686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,51200,0.7297342088487414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,12288,0.050698667764663696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,16384,0.23573867479960123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,10240,0.0422017772992452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,8192,0.032840887705485024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,12288,0.17437067296769884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,7168,0.02796088986926609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,10240,0.1432551145553589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,6144,0.024122666981485155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,7168,0.09982311063342625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,5120,0.020948444803555805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,6144,0.08645777569876777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,4096,0.01755555636352963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,5120,0.07348089085684882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,4096,0.059486223591698535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,3584,0.016255110502243042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,3072,0.01423111061255137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,3584,0.052601777844958834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,2560,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,3072,0.04606666829850939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,2048,0.010405333505736457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,2560,0.03983822133806016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,1536,0.00870311094654931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,2048,0.033646222617891096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,1024,0.0069982219073507525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,8192,0.11352177460988362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,1536,0.02730400032467312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,768,0.006953777538405524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,1024,0.021631111701329548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,512,0.006063999815119638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,768,0.01869066721863217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,256,0.004973333328962326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,512,0.016238222519556682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,128,0.0043333334227403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,256,0.014466666513019137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,64,0.004045333299371931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,128,0.014142221874660917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,2048,32,0.004015111260943943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,64,0.013743999931547375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,2048,32,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,65536,0.16769866148630777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,51200,0.13163467248280844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,16384,0.04488977789878845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,16384,0.21796354982588026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,12288,0.03452444407675002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,65536,0.8390337626139323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,10240,0.03099200129508972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,51200,0.6634115642971462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,12288,0.1615697807735867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,8192,0.025038222471872967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,7168,0.02239022155602773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,10240,0.13197244538201228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,6144,0.02051111062367757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,8192,0.1045244468583001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,5120,0.0166640000210868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,6144,0.07932888799243502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,4096,0.01440177857875824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,7168,0.09273155530293782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,5120,0.06684711244371203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,3584,0.01315111087428199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,3072,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,4096,0.0539591113726298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,2560,0.010084444450007545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,3584,0.047953777843051486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,2048,0.008821333448092142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,2560,0.03607555561595493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,1536,0.008685333033402761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,3072,0.042211555772357516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,2048,0.030209776428010728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,1024,0.006559111177921295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,768,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,1536,0.02509333358870612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,1024,0.020261334048377145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,512,0.004960000101062986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,256,0.00425511101881663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,768,0.017599110802014668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,512,0.015536889433860779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,256,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,128,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,64,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1536,32,0.003970666478077571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,128,0.013436444103717804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,64,0.013457777599493662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1536,32,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,65536,0.1620311074786716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,51200,0.12966488467322454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,16384,0.04368622104326884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,16384,0.20672710736592612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,12288,0.03199199835459391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,51200,0.6232497957017686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,65536,0.794156445397271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,10240,0.0273324449857076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,12288,0.15277600288391113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,8192,0.02165066699186961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,7168,0.01956888867749108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,10240,0.12486933337317573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,6144,0.01750577820671929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,8192,0.09821510977215236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,5120,0.01510933372709486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,6144,0.07548533545600043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,4096,0.012492444780137805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,4096,0.050776001479890615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,3584,0.011194667054547204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,7168,0.08584888776143391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,3072,0.010044444766309526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,3584,0.04494577646255493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,2560,0.008776889079146916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,3072,0.03890577620930142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,2048,0.007720888488822513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,5120,0.06302044126722547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,1536,0.006798222247097228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,2560,0.03382133444150289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,2048,0.02844266759024726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,1536,0.02398577829202016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,768,0.0053591111467944244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,1024,0.01926933394538032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,512,0.0046471113132105935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,768,0.016919111212094624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,256,0.004001777619123459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,512,0.014847111370828418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,128,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,256,0.013183111117945777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,128,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,64,0.003491555651028951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,32,0.00349866681628757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,64,0.012515555653307172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,1024,1024,0.005692444327804778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,1024,32,0.012503110700183444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,65536,0.1189919975068834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,51200,0.09422755241394043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,16384,0.03266400098800659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,16384,0.19681244426303438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,51200,0.5918213526407877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,12288,0.023360000716315374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,10240,0.02023555503951179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,65536,0.7562479972839355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,8192,0.01688088807794783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,12288,0.1459395620557997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,10240,0.11884444289737278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,8192,0.0935937762260437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,6144,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,6144,0.07200533151626587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,5120,0.01181244436237547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,5120,0.06073421902126736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,7168,0.01516266663869222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,4096,0.010056888891590966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,3584,0.009071110851234859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,4096,0.047931555244657725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,3072,0.00832177781396442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,3584,0.042734222279654614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,2560,0.0074720001882976955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,3072,0.03732000125779046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,2048,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,2560,0.032650666104422674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,2048,0.0277804434299469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,1536,0.005693333430422678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,1024,0.00527555536892679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,1536,0.023051554958025616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,768,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,1024,0.01888177792231242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,768,0.016314667132165697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,512,0.004290666845109728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,7168,0.08184178007973565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,256,0.003630222131808599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,128,0.0032604444358083936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,512,0.014560888210932413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,256,0.012819555898507437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,64,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,768,32,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,128,0.012598222328556908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,64,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,768,32,0.011794666449228922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,65536,0.09505510992474026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,51200,0.07858044571346708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,16384,0.030561778280470107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,16384,0.19534222284952799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,12288,0.024863110648261175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,65536,0.7479182349310981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,10240,0.021720889541837905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,51200,0.5885190963745117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,8192,0.0185244447655148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,12288,0.14495200581020778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,7168,0.0158942225906584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,10240,0.11833688947889541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,6144,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,5120,0.009696889254781935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,7168,0.08110577530331083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,4096,0.009724444813198512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,5120,0.05921155876583523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,6144,0.07079555590947469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,3584,0.009219555391205682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,4096,0.04654666781425476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,3072,0.009152889251708984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,3584,0.04162133402294583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,2560,0.007621333003044128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,2560,0.03197955422931247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,2048,0.006680000159475539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,2048,0.027112000518374976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,8192,0.09322222073872884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,1536,0.005921777751710679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,1024,0.004976888911591636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,1536,0.023021333747439917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,1024,0.018445332845052082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,768,0.004569777597983678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,512,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,3072,0.03773155477311876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,768,0.015861334072219003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,256,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,128,0.003261333331465721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,256,0.012470222181744046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,64,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,512,0.014599111345079211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,128,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,512,32,0.0035386664999855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,64,0.01180622229973475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,512,32,0.011714666253990598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,65536,0.04941155513127645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,51200,0.041208889749315046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,16384,0.01640888883007897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,16384,0.18617244561513266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,12288,0.013857777747843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,65536,0.7101999918619791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,51200,0.5582577917310927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,10240,0.012701333396964602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,8192,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,12288,0.13773689005109999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,7168,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,8192,0.08736177947786118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,10240,0.11226488484276666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,6144,0.010021333065297868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,7168,0.07670044236712985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,5120,0.009008888569143083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,4096,0.007615110940403408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,5120,0.056221332814958364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,3584,0.0069395556218094295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,4096,0.04418933391571045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,3072,0.006299555715587404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,3584,0.03980977667702569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,2560,0.0058017778727743365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,3072,0.034288889831966825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,2048,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,2560,0.030282666285832722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,1536,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,2048,0.025715554753939312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,1024,0.0041982221106688184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,1536,0.02176711128817664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,768,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,1024,0.017874666386180453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,512,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,768,0.01551644504070282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,6144,0.06707021925184461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,256,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,512,0.013501333693663279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,128,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,64,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,128,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,256,0.012472000386979846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,64,0.01148088855875863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,65536,0.0435804459783766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,256,32,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,51200,0.03522933191723294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,65536,0.7094240188598633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,16384,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,51200,0.5578293270534939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,256,32,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,12288,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,16384,0.18585422303941515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,10240,0.009752888646390703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,8192,0.009007111191749573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,12288,0.1379128826989068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,7168,0.008655110994974772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,10240,0.11149688561757405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,6144,0.008041777544551426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,7168,0.07633599970075819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,5120,0.008921778036488427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,6144,0.06616889105902778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,4096,0.007366221812036302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,5120,0.055045333173539907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,4096,0.04288444585270352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,3584,0.039581331941816546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,3072,0.006335999816656113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,3072,0.034745779302385114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,2560,0.005994666781690385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,8192,0.0876035557852851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,2048,0.005434666656785541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,2560,0.030101332399580214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,1536,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,1024,0.004287999951177173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,1536,0.021898667017618816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,2048,0.026072000463803608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,768,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,1024,0.01789155602455139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,512,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,768,0.015495111544926962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,256,0.0033093332830402586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,512,0.013830222189426422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,128,0.003173333282272021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,256,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,64,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,128,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,32,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,64,0.01127022256453832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,65536,0.03818044397566054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,128,32,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,51200,0.03251111176278856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,128,3584,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,16384,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,16384,0.18559022744496664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,65536,0.7088489002651639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,51200,0.5571280055575901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,10240,0.009376000199053023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,12288,0.13723022407955593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,8192,0.009731555150614845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,10240,0.11155554983350967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,7168,0.009305777649084726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,8192,0.08733866612116496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,6144,0.008637333081828224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,7168,0.07664444711473253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,5120,0.007968000239796108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,6144,0.06621866756015353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,4096,0.007314666277832455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,5120,0.05481600099139743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,3584,0.006589333216349284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,12288,0.010020444790522257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,3072,0.006255110932721033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,4096,0.043706665436426796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,2560,0.005559111220969095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,3584,0.03891822364595201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,3072,0.03410044312477112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,2560,0.03018222252527873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,1536,0.004554666578769684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,2048,0.02573066618707445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,1024,0.003949333396222857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,1536,0.02160177793767717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,768,0.0035822221802340615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,1024,0.01755555636352963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,512,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,768,0.015469332536061605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,256,0.0031840000301599503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,512,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,128,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,256,0.01204355557759603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,64,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,128,0.01145066652033064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,32,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,64,2048,0.00488711107108328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,64,0.01126844435930252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,65536,0.03674577673276266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,51200,0.029628445704778034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,64,32,0.01128000020980835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,16384,0.0141004439857271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,16384,0.18545866012573242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,12288,0.01514844430817498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,65536,0.7088666492038302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,51200,0.556895997789171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,10240,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,8192,0.011735111474990845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,12288,0.13750221994188097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,7168,0.011032000184059143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,10240,0.1116773287455241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,6144,0.00999111102686988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,8192,0.08708088927798802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,5120,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,7168,0.07611199882295397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,4096,0.008302222523424361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,6144,0.06530221965577868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,3584,0.0076248885856734375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,5120,0.05502400133344862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,3072,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,4096,0.043828444348441235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,2560,0.006256000035338932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,3584,0.03905066516664293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,2048,0.005555555638339784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,3072,0.03461511267556085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,1536,0.004538666870858935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,2560,0.030051555898454454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,1024,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,2048,0.025702221526039973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,768,0.0034862222770849862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,1536,0.021713778376579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,512,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,1024,0.01756888959142897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,256,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,768,0.015528000063366361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,512,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,64,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,256,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1537,32,32,0.0027991111079851785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,64,0.011214222345087262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,32,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1537,32,128,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,51200,4.728973388671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,16384,1.5530302259657118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,51200,7.2773225572374125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,16384,2.1125848558213973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,12288,1.1494400236341689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,12288,1.501546647813585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,16384,1.5922799640231664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,51200,5.039748297797309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,10240,0.956608878241645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,10240,1.229466650221083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,12288,1.2050249311659071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,8192,0.7757831149631076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,7168,0.6573208702935113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,10240,1.0076995425754123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,8192,0.9672844145033094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,6144,0.568336009979248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,8192,0.813384903801812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,7168,0.8406355645921496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,7168,0.7171200116475424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,5120,0.47919379340277773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,6144,0.7168551021152072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,4096,0.38082223468356663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,5120,0.5967822074890137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,6144,0.6225866741604275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,5120,0.5266586409674751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,3584,0.3356835577223036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,4096,0.47744178771972656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,4096,0.43304888407389325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,3584,0.41652266184488934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,3072,0.30927732255723744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,3584,0.38596532079908585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,3072,0.36128711700439453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,2560,0.2512213389078776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,3072,0.3391039901309543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,2048,0.20889155069986978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,2560,0.3110462294684516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,1536,0.1442835595872667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,2560,0.2906373341878255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,2048,0.2551315625508626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,2048,0.24424354235331217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,1536,0.20165244738260904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,1536,0.1972479952706231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,768,0.08809155225753784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,1024,0.15467022524939641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,1024,0.15399821599324545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,512,0.06210933128992716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,768,0.12358578046162923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,768,0.1289893388748169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,256,0.048319111267725624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,512,0.10608088970184326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,512,0.10611377822028266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,128,0.042845331960254245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,256,0.0859911110666063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,1024,0.10807022121217515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,128,0.09015822410583496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,64,0.03603377938270569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,65536,32,0.03821244504716661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,64,0.09023200141059028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,32,0.0902071131600274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,65536,256,0.09567289219962226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,65536,128,0.06938133637110393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,65536,4.587572309705946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,51200,3.515103022257487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,65536,7.332881503634983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,16384,1.1483501858181424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,51200,5.75359386867947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,16384,1.6595431433783636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,65536,5.299354553222656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,12288,0.8483689096238879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,12288,1.201216909620497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,16384,1.3065182367960613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,10240,0.740787559085422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,51200,4.09644529554579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,10240,0.9771173265245227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,8192,0.5705351299709743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,8192,0.7667804294162326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,10240,0.8195706473456489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,7168,0.48371023601955837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,12288,0.9808133443196615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,8192,0.6635991202460395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,7168,0.6782551341586642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,6144,0.42528888914320206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,5120,0.35686222712198895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,7168,0.5843991173638238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,6144,0.5709875424702963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,4096,0.2875422106848823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,6144,0.5083635648091634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,5120,0.49014843834771055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,4096,0.3822302288479275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,5120,0.43123555183410645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,3584,0.26322844293382436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,4096,0.35569156540764707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,3072,0.2288604312472873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,3584,0.33479823006523984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,2560,0.17880533801184761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,3072,0.291648891237047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,3072,0.31598488489786786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,3584,0.315940433078342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,2048,0.16215556197696263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,2560,0.24011733796861437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,2560,0.2508640024397108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,1536,0.11197066307067871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,2048,0.20591644446055093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,2048,0.20306666692097983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,1536,0.16332088576422796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,1024,0.08345511224534775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,1536,0.16373244921366373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,768,0.07000977463192411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,1024,0.12509332762824163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,1024,0.12897688812679714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,768,0.10016089015536839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,512,0.04920088913705614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,768,0.10759111245473225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,256,0.03740533192952474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,512,0.08611644638909234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,512,0.08988000286950006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,128,0.032608889871173434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,256,0.07288000318739149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,256,0.07744889126883613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,64,0.02778844369782342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,128,0.07387999693552653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,51200,32,0.028562668297025893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,64,0.07330666648017035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,51200,32,0.0739333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,51200,128,0.06018400192260742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,65536,1.5522204505072699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,51200,1.2181350919935439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,65536,2.6938578287760415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,51200,2.1063609653049045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,16384,0.3933600054846869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,51200,1.8706915113661025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,65536,2.390390184190538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,16384,0.6286408636305068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,12288,0.3254355589548747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,12288,0.4544826613532172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,10240,0.24931645393371582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,16384,0.6154426468743218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,12288,0.46145953072441953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,8192,0.21873154905107287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,10240,0.37288355827331543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,8192,0.295629342397054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,7168,0.17831822236378989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,10240,0.3858328925238715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,8192,0.31091467539469403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,6144,0.16113689210679796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,7168,0.26034045219421387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,5120,0.14145599471198186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,7168,0.2757413387298584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,6144,0.22426933712429473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,6144,0.24066223038567436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,4096,0.11324621571434869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,5120,0.18834843900468615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,3584,0.08635199732250637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,5120,0.2052373356289334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,4096,0.1541422208150228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,3584,0.1367573340733846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,3584,0.15247733063167995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,3072,0.08549333280987209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,2560,0.06905333201090495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,3072,0.13596444659762913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,3072,0.11976710955301921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,2048,0.05875111288494534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,2560,0.10429422060648601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,2560,0.11810844474368626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,1536,0.04420800010363261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,2048,0.08642133076985677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,2048,0.10149511363771226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,1024,0.030889779329299927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,1536,0.08337155315611099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,1536,0.06894399722417195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,4096,0.16973422633277047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,768,0.023751111494170293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,1024,0.06676089101367526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,1024,0.05368622144063314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,512,0.01886399918132358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,768,0.04247911108864678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,768,0.05705600314670139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,256,0.014832889040311178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,512,0.049351112710105054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,512,0.03593422306908502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,256,0.03089955449104309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,128,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,256,0.04150755537880792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,64,0.014550222290886773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,16384,32,0.014845333165592618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,16384,128,0.03309333324432373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,64,0.02982399861017863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,32,0.029209779368506536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,65536,1.0905084609985352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,16384,128,0.029887109994888306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,51200,0.8686541981167264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,65536,2.1645083957248263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,16384,0.2915066613091363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,51200,1.6879466374715169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,16384,0.5050382084316677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,65536,2.0048906538221574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,12288,0.22207732995351157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,12288,0.37237244182162815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,16384,0.5188186433580186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,12288,0.38924800025092227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,10240,0.17844533920288086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,10240,0.304550223880344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,51200,1.5717048645019531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,8192,0.1472017765045166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,10240,0.32474133703443736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,7168,0.1286417775683933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,8192,0.24056267738342285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,8192,0.2612977822621663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,6144,0.11183555920918782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,7168,0.21310310893588594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,7168,0.23228089014689127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,5120,0.09903199805153741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,6144,0.18360088931189644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,6144,0.20309600565168592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,4096,0.0746142201953464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,5120,0.15517511632707384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,5120,0.17300977971818712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,3584,0.06811822122997709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,4096,0.12751466698116726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,4096,0.14463199509514704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,3072,0.06191821893056234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,3584,0.1135831144120958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,3584,0.13017511367797852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,3072,0.10003111097547744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,3072,0.11559289031558567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,2560,0.049448887507120766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,2048,0.04083911246723599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,2560,0.1013066636191474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,2048,0.07230489121543036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,1536,0.031847112708621554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,1536,0.05716800027423435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,1536,0.0729093352953593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,1024,0.02312533391846551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,2560,0.08682488732867771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,2048,0.08709066443973118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,1024,0.059042665693495006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,768,0.018215111560291715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,1024,0.04423822296990288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,512,0.01388177772363027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,768,0.03571644425392151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,768,0.05047288868162367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,256,0.011104889214038849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,512,0.030419554975297716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,512,0.043613334496816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,128,0.010441777606805166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,256,0.02608266638384925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,256,0.03760266635153029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,12288,128,0.03017866611480713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,128,0.025091555383470323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,32,0.009310222334331935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,64,0.025111999776628282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,12288,32,0.02472355630662706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,12288,64,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,65536,1.0166506237453885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,51200,0.7964933183458117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,65536,1.917692396375868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,51200,1.4808808432685003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,16384,0.2665004465315077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,65536,1.8710933261447484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,16384,0.4494489034016927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,12288,0.20194221867455375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,51200,1.468306647406684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,10240,0.1680124468273587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,12288,0.3305937714046902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,16384,0.48459021250406903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,8192,0.13268800576527914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,12288,0.364865779876709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,10240,0.26978932486640084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,10240,0.30271556642320424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,8192,0.2145084407594469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,7168,0.1252115567525228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,6144,0.09735021988550822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,8192,0.24560978677537706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,7168,0.18846133020189074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,7168,0.21753333674536812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,5120,0.08782044384214614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,6144,0.16311555438571507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,6144,0.18977067205641004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,4096,0.06601955493291219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,5120,0.13779732916090223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,5120,0.16258845064375135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,3584,0.060608002874586314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,4096,0.1145733329984877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,4096,0.1358026663462321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,3584,0.1021617783440484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,3072,0.054175111982557506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,3584,0.12151200241512722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,2560,0.045596443944507174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,3072,0.08977422449323867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,3072,0.10896355575985378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,2048,0.03521155648761325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,2560,0.07806489202711317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,2560,0.09548622369766235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,2048,0.06407288710276286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,2048,0.08240444130367704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,1024,0.02019644445843167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,1536,0.05113688773579068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,1024,0.04003377755482992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,1024,0.05599911345375908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,768,0.016908443636364408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,768,0.03239911132388645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,1536,0.027220444546805486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,768,0.04841599861780802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,512,0.014783999986118741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,1536,0.06771466467115614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,512,0.02750933335887061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,256,0.02407111061943902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,256,0.03508177730772231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,128,0.010775110787815519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,128,0.023036445180575054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,512,0.0413626664214664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,256,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,10240,128,0.030418667528364394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,64,0.022707555029127333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,32,0.009482666850090027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,10240,32,0.02239111065864563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,65536,0.8243190977308484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,10240,64,0.00963466697269016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,51200,0.6394293573167589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,65536,1.6482071346706813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,51200,1.2853084140353732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,16384,0.21444177627563477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,65536,1.7400862375895183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,16384,0.39176000489128965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,51200,1.3667866388956706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,12288,0.16277955638037786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,16384,0.45267465379503036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,12288,0.28818845748901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,10240,0.1336115598678589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,12288,0.3391973442501492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,8192,0.10995999972025554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,10240,0.2359671062893338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,10240,0.28248800171746147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,8192,0.18753955099317762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,8192,0.22759199142456055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,7168,0.09544799725214641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,6144,0.08172977632946439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,7168,0.16528533564673528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,7168,0.20134577486250135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,5120,0.07288444704479642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,6144,0.14357154899173313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,6144,0.17643377516004774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,4096,0.057144886917538114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,5120,0.12112443976932102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,5120,0.15174933274586996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,4096,0.09988888767030503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,4096,0.12671021620432535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,3584,0.05008622341685825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,3072,0.044109334548314415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,3584,0.11331021785736084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,3584,0.09049155314763387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,2560,0.034479998879962497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,3072,0.07983733548058404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,3072,0.10186133119795059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,2048,0.0314044455687205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,2560,0.06844000021616618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,2560,0.08933155404196845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,1536,0.023440000083711412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,2048,0.0765591131316291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,2048,0.05639288822809855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,1024,0.016922665966881644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,1536,0.06406133042441474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,1536,0.0443795555167728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,1024,0.05223377876811557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,768,0.014775110615624322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,512,0.01223111069864697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,768,0.04493777619467842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,768,0.028287112712860107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,256,0.009783999787436591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,512,0.024384000235133704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,512,0.03870755434036255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,128,0.009882666998439366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,256,0.021329777108298406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,256,0.03277511066860623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,64,0.00964355551534229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,128,0.02066044509410858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,8192,32,0.009673777553770278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,8192,128,0.027973333994547527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,64,0.020693333612547982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,32,0.020259555843141343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,8192,1024,0.03497599893146091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,65536,0.6519973542955186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,51200,0.5670080184936523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,65536,1.5238222546047633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,51200,1.1888550652398002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,16384,0.1666897800233629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,16384,0.36546754837036133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,65536,1.6123733520507812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,12288,0.12925599681006536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,16384,0.4210417800479465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,51200,1.2661369111802843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,10240,0.10906577772564358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,12288,0.2694106631808811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,8192,0.0840453306833903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,10240,0.22037955125172934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,12288,0.3156195481618245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,7168,0.08194577693939209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,10240,0.26316356658935547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,8192,0.17408622635735405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,8192,0.21153067217932808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,6144,0.07091110944747925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,7168,0.15405421786838108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,7168,0.18732178211212158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,5120,0.057462222046322294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,6144,0.13368622461954752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,6144,0.1647404432296753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,4096,0.04782399866316053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,5120,0.11316800117492676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,5120,0.1411155594719781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,3584,0.04227733280923632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,4096,0.09284622139400905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,4096,0.11800532870822483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,3072,0.036349332994884916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,3584,0.08156622118420072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,3584,0.10619022448857625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,2560,0.031347556246651545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,3072,0.07392711109585233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,3072,0.09554133150312637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,2048,0.025587555434968736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,2560,0.06269066863589816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,2560,0.08324266804589166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,1536,0.02016088863213857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,2048,0.05232000019815233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,2048,0.0724195573065016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,1024,0.015007111761305066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,1536,0.041590223709742226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,1024,0.03264266583654616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,768,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,1024,0.04974400003751119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,768,0.026502221822738647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,768,0.042805333932240806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,512,0.009740444521109263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,512,0.02306755549377865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,512,0.03583822316593594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,256,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,256,0.01999288962946998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,256,0.030945777893066406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,128,0.006913777854707506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,1536,0.060257779227362744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,128,0.0196106665664249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,7168,128,0.027690667245123122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,32,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,64,0.019215111931165058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,7168,32,0.018938667244381376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,65536,0.5439110861884223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,7168,64,0.006255110932721033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,51200,0.4648142390780979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,65536,1.3960000144110785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,51200,1.1040222379896376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,65536,1.484206199645996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,16384,0.14799643887413874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,16384,0.33624709977044004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,51200,1.1658088896009657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,12288,0.10876088672214085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,16384,0.3884906768798828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,10240,0.09757066435284084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,12288,0.2914515601264106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,10240,0.20361244678497314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,8192,0.08121955394744873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,10240,0.24157688352796766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,8192,0.16085599528418645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,8192,0.19581333796183267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,7168,0.07120800018310547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,12288,0.2495511107974582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,7168,0.14238844977484807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,7168,0.17334132724338105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,6144,0.05995021926032173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,5120,0.05086044470469157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,6144,0.12355911731719971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,6144,0.15200266573164198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,4096,0.041788445578681103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,5120,0.10532622204886542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,5120,0.13060888979170057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,3584,0.0369102226363288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,4096,0.11004533370335896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,3584,0.07650133636262682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,3584,0.09848178095287746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,3072,0.0316826668050554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,3072,0.08824977609846328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,2560,0.026664889521068994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,2560,0.059192001819610596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,2048,0.02237333357334137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,2560,0.0779822203848097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,2048,0.0486195551024543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,1536,0.017921777235137094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,2048,0.0683271090189616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,1536,0.03877777854601542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,1024,0.013610666824711693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,1536,0.05713866816626655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,1024,0.03104089034928216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,3072,0.06606311268276639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,1024,0.047325332959493004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,768,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,512,0.008646222452322641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,768,0.025133333272404138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,512,0.021292444732454088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,256,0.006671999891599019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,768,0.040992889139387347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,512,0.03365155392222934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,4096,0.0866960022184584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,256,0.018920888503392536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,128,0.006592888798978593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,256,0.030132446024152968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,64,0.005991111199061076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,6144,128,0.02676888969209459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,128,0.018582221534517076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,6144,32,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,64,0.018251554833518136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,6144,32,0.018283555905024212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,65536,0.5362728966606988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,51200,0.435440010494656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,65536,1.2624390920003254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,51200,0.9979475869072808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,65536,1.4830737643771703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,16384,0.13753333356645372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,16384,0.3111786577436659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,12288,0.10108000040054321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,16384,0.38761244879828555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,51200,1.1644586986965604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,10240,0.09474400016996597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,12288,0.22959287961324057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,12288,0.29166221618652344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,8192,0.07017955515119764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,10240,0.18814755810631645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,10240,0.24113689528571236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,8192,0.14867199791802302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,8192,0.1952702204386393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,7168,0.13127377298143175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,6144,0.052447110414505005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,7168,0.17316799693637422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,6144,0.11441155274709065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,6144,0.15139467186397976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,5120,0.04411022199524773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,5120,0.09727466768688625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,4096,0.0352151095867157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,7168,0.06150044335259331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,4096,0.07956000169118245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,4096,0.10913866758346558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,3584,0.06994666655858357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,3584,0.09782666630215114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,3072,0.028096887800428603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,5120,0.13013777467939588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,3072,0.0627688897980584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,2560,0.022720000810093347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,3072,0.08803288804160224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,2560,0.05204000075658163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,3584,0.03204977843496535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,2048,0.018943111101786297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,2560,0.07659022013346355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,1536,0.01608533329433865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,2048,0.04437866806983948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,2048,0.0671608911620246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,1024,0.012085333466529846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,1536,0.035820444424947105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,1536,0.056391113334231906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,768,0.010488000180986194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,1024,0.028655999236636694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,1024,0.046253333489100136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,512,0.00868088917599784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,768,0.023390221926901076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,768,0.039048890272776283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,256,0.007272889216740926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,512,0.019938665959570143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,512,0.03295733200179206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,128,0.006509333435032103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,256,0.017872888180944655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,64,0.007610666255156199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,128,0.025980444418059453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,5120,32,0.007471111085679796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,64,0.01716444392999013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,128,0.017616889543003507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,5120,32,0.016813332835833233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,65536,0.4090782271491156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,51200,0.32170844078063965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,5120,256,0.02843377656406826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,65536,1.1538000106811523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,16384,0.11065422164069282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,51200,0.9011946784125434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,65536,1.3562569088406031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,12288,0.08532000250286526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,16384,0.3554657830132379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,51200,1.0646115409003365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,16384,0.2845875687069363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,12288,0.21061510509914824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,12288,0.2679022153218587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,10240,0.1731475591659546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,8192,0.05607199668884277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,8192,0.17820889419979521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,7168,0.049182222949133984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,8192,0.13744621806674534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,7168,0.1208675569958157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,10240,0.07456800010469225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,6144,0.10484088791741265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,10240,0.22163377867804634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,7168,0.15919110510084364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,5120,0.03624799847602844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,6144,0.13935289118025038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,4096,0.02979644470744663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,5120,0.1203120019700792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,5120,0.09062133232752483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,4096,0.07468089130189684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,3584,0.026403556267420452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,6144,0.04163377814822727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,3584,0.06620888577567206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,3584,0.09049689107471043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,4096,0.10125066836675008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,3072,0.05685066514545017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,2560,0.01978133287694719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,3072,0.08207022481494479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,2560,0.048384000857671104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,2048,0.016928889685206942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,2560,0.07223111391067505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,3072,0.023208000593715247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,1536,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,2048,0.06300533480114408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,1536,0.03361422154638503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,1024,0.010432889064153036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,1536,0.05309333403905233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,1024,0.026165333059098985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,768,0.008711111214425828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,1024,0.04191199938456217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,768,0.0219217770629459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,512,0.007739555504586961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,768,0.03770400087038676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,512,0.018664888209766813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,256,0.006768888897365994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,512,0.03200533323817783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,256,0.01660977800687154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,128,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,128,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,256,0.028975112570656672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,64,0.005976000179847081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,2048,0.04045866595374213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,4096,128,0.024812445044517517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,4096,32,0.0063368889192740125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,32,0.01553244392077128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,4096,64,0.01612622208065457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,65536,0.3597448931799994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,51200,0.2778595553504096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,65536,1.0742977990044487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,51200,0.8466417524549695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,16384,0.09810577498541938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,65536,1.356124454074436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,16384,0.2700124316745334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,12288,0.07252799802356295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,12288,0.19909600416819254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,16384,0.35666312111748594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,51200,1.065400865342882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,10240,0.06453688939412434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,10240,0.16355289353264704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,12288,0.2674070994059245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,8192,0.05035110976960924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,10240,0.2217937840355767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,8192,0.12924177116817898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,8192,0.17875377337137857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,7168,0.11398311456044515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,7168,0.1588746706644694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,6144,0.035785777701271906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,6144,0.09982133573955959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,6144,0.1390062173207601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,5120,0.03058755397796631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,5120,0.08486933178371853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,5120,0.12033155229356553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,4096,0.02482755482196808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,7168,0.04372622238265144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,4096,0.07022311290105183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,3584,0.023253333237436082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,4096,0.10054933362536961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,3072,0.020449777444203694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,3584,0.06199644671546089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,2560,0.017855111095640395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,3072,0.05371110969119602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,3584,0.09068533447053696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,2048,0.015090665883488126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,2560,0.04522044459978739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,2560,0.07105244530571832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,1536,0.01200888885392083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,2048,0.06246755520502726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,1536,0.031594667169782854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,2048,0.038979556825425886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,1024,0.00906844437122345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,1536,0.051968889103995435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,768,0.008099555969238281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,1024,0.041103111373053655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,1024,0.02466222147146861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,768,0.020982222424613103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,512,0.006950221955776215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,768,0.035980443159739174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,256,0.005943111247486538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,512,0.01790488925245073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,256,0.016168889072206285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,3072,0.08168533113267687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,128,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,256,0.027815110153622095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,64,0.004941333499219683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,128,0.015527110960748462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,128,0.024714666936132643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3584,32,0.00490133340160052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,64,0.015473778049151102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3584,32,0.015394666128688388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,65536,0.2934177716573079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3584,512,0.03162044286727905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,65536,1.01946226755778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,51200,0.24462845590379503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,65536,1.2284666697184246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,16384,0.08376977841059367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,51200,0.8032791349622938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,16384,0.2581884331173367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,51200,0.9659609264797635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,12288,0.0594924423429701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,16384,0.3237137794494629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,10240,0.05089777708053589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,12288,0.24301332897610137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,12288,0.19086578157213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,10240,0.15610133277045354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,8192,0.04095822241571214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,10240,0.2013546625773112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,8192,0.12358933024936253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,7168,0.03748177819781833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,8192,0.16254399882422552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,7168,0.10905155870649551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,6144,0.032764444748560585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,7168,0.1448462274339464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,5120,0.027830223242441814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,6144,0.09513511260350545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,5120,0.08135110802120633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,6144,0.1274462276034885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,4096,0.022533333963818018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,5120,0.10976177453994751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,3584,0.02032444377740224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,4096,0.09242755836910671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,4096,0.06712177726957533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,3584,0.05864444706175062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,3584,0.08357777860429551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,3072,0.05106311043103536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,3072,0.07490311066309611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,2560,0.0158897770775689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,2560,0.04324088825119866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,2560,0.06587022542953491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,2048,0.013887999786270989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,2048,0.036764442920684814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,1536,0.01131199962562985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,1536,0.030227555168999568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,3072,0.018363555272420246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,1024,0.008748444418112436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,1536,0.04849422309133741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,1024,0.03876711262596978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,1024,0.023677332533730402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,768,0.007649777664078607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,768,0.03396711084577773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,2048,0.05786844756868151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,512,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,768,0.02028622229894002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,256,0.0053004444473319584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,512,0.03016177813212077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,256,0.02713866697417365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,256,0.01595288846227858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,128,0.004563555535342959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,3072,128,0.023519999451107446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,64,0.004272888931963178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,128,0.015233778291278414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,3072,32,0.004261333495378494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,64,0.014803555276658801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,32,0.014872888724009195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,65536,0.25832443767123753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,3072,512,0.017298666967286002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,51200,0.21532532903883192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,65536,0.9735244115193685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,51200,0.7603973282708062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,16384,0.0726977785428365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,65536,1.2266239590115018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,16384,0.24504266844855416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,12288,0.05505688985188802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,16384,0.32418399386935765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,10240,0.04722933305634392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,12288,0.18183999591403535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,12288,0.24308178159925672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,51200,0.9639626608954536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,10240,0.20111555523342559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,10240,0.14997422695159912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,8192,0.11852444542778863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,7168,0.03356711069742838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,8192,0.16219288773006862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,6144,0.030135111676322088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,7168,0.10430310832129584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,7168,0.1445466677347819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,5120,0.026054221722814772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,6144,0.09130311012268066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,6144,0.12638400660620794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,4096,0.021939555803934734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,5120,0.10952888594733344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,8192,0.03834044602182176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,5120,0.07746755414538913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,3584,0.01962844365172916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,4096,0.06302755408816867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,4096,0.0920799970626831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,3072,0.017298666967286002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,3584,0.08237866560618083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,3584,0.055580443806118436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,3072,0.07487999730639987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,3072,0.04864177770084805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,2560,0.015512888630231222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,2048,0.013197333448463015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,2560,0.06589511368009779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,2560,0.04153866569201151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,1536,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,2048,0.057238221168518066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,2048,0.03535111082924737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,1024,0.00871377769443724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,1536,0.028408888313505385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,1536,0.04866488774617513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,768,0.007628444168302748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,1024,0.038458665211995445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,1024,0.023038221730126276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,512,0.006592888798978593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,768,0.019344889455371432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,768,0.033667554457982383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,256,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,512,0.02947910957866245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,512,0.017068444026841056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,256,0.014821334017647637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,128,0.005072000126043956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,256,0.026505778233210247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,64,0.004868444469239977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,128,0.014477334088749357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2560,32,0.006264888991912206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,64,0.014172444740931192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2560,128,0.02356622285313076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2560,32,0.013797333670987023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,65536,0.21254400412241617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,51200,0.1495128870010376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,65536,0.922393798828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,51200,0.7212079895867242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,16384,0.05770844221115112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,65536,1.2258408864339192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,16384,0.2338177892896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,12288,0.041069332096311785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,16384,0.32375023100111217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,12288,0.1734586689207289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,51200,0.9646399815877279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,10240,0.035275555319256253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,12288,0.2427297698126899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,8192,0.030029333300060693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,10240,0.20112710528903535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,10240,0.1422035561667548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,7168,0.025807999902301367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,8192,0.11236000061035156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,8192,0.16222222646077475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,6144,0.023754666248957317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,7168,0.14386133352915445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,7168,0.09946222437752618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,6144,0.08658399846818711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,6144,0.12599378161960179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,5120,0.07323733303281996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,5120,0.10894666777716742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,4096,0.01701866587003072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,4096,0.059260441197289355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,3584,0.01534577707449595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,3584,0.05204977922969394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,3584,0.08286755614810519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,5120,0.020370667179425556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,3072,0.014010666145218743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,3072,0.07484444644716051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,2560,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,4096,0.09154399898317124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,2560,0.03911822372012668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,2048,0.010266666611035665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,2560,0.06525778108172946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,2048,0.03293066554599338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,1536,0.008740444150235917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,2048,0.05518488751517402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,1536,0.027465777264700994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,1536,0.04626399940914578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,1024,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,1024,0.021663111117151048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,1024,0.038020445240868464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,3072,0.04639644424120585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,768,0.006301333506902059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,512,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,768,0.018574222922325134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,512,0.016224000189039443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,256,0.004592888885074192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,768,0.033746666378445096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,512,0.029493331909179688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,128,0.004382222063011593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,256,0.014479110638300577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,128,0.013787555197874704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,256,0.026885333988401625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,64,0.004179555508825514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,2048,128,0.02347822156217363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,2048,32,0.0041164445380369825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,65536,0.1732871135075887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,32,0.013463999662134381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,2048,64,0.013847110999955071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,51200,0.12711644172668457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,65536,0.8412453333536783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,16384,0.04704977737532722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,51200,0.6637484232584635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,65536,1.0988044738769531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,16384,0.21678845087687174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,12288,0.03268622358640035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,51200,0.8636213408576118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,16384,0.29122667842441136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,12288,0.16091644763946533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,12288,0.2189075549443563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,10240,0.130986664030287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,8192,0.022424888279702928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,10240,0.18025955888960096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,8192,0.10378311077753703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,8192,0.14550222290886775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,7168,0.09131911065843369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,10240,0.02807377775510152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,6144,0.01889333294497596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,7168,0.1290417777167426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,6144,0.07967378033532037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,6144,0.11415644486745198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,5120,0.015860444969601102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,5120,0.06781422429614596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,5120,0.09867822461658055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,4096,0.054173333777321704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,4096,0.08321599827872382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,3584,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,3584,0.04831733306248983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,3584,0.07509244150585599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,3072,0.010910222099887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,4096,0.01311644415060679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,3072,0.042282667424943715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,3072,0.0677128897772895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,2560,0.009840889109505547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,2560,0.05949155489603678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,2560,0.036244445376926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,2048,0.00832799987660514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,2048,0.03056710958480835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,2048,0.05074310965008206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,1536,0.007080888582600488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,1536,0.025451555848121643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,1536,0.04269511169857449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,1024,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,7168,0.020516445239384968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,768,0.005214222189452913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,1024,0.03547822104560004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,768,0.01722044414944119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,512,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,512,0.015289778510729471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,512,0.027823110421498615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,256,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,256,0.013760888742076026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,1024,0.02057422200838725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,256,0.025359110699759588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,768,0.031207998593648274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,128,0.013417777915795645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,64,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1536,128,0.02264533274703556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,32,0.0038488888078265717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,64,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,65536,0.13204356034596762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1536,128,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,65536,0.7909466425577799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,51200,0.10641688770718044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1536,32,0.013061333033773633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,16384,0.03706489006678263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,16384,0.2062764432695177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,65536,1.0953590604994032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,12288,0.027128888501061335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,16384,0.29042310184902614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,12288,0.15281866656409368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,10240,0.021680000755521987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,51200,0.8624924553765191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,12288,0.2180915541119046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,8192,0.01904444396495819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,10240,0.12433244122399224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,10240,0.179967999458313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,7168,0.01667644414636824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,8192,0.09798489014307658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,51200,0.6213137838575575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,8192,0.14462933275434706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,6144,0.01479111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,7168,0.12896622551812068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,7168,0.08654577864540948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,6144,0.07464444637298584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,5120,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,6144,0.1128115521536933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,4096,0.011344888971911537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,5120,0.09738488992055257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,5120,0.06304622358746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,4096,0.05039466751946343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,4096,0.08191821972529094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,3584,0.009767110976907942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,3072,0.009012444151772393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,3584,0.044821331898371376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,3072,0.03907733493381076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,3072,0.06740888622072008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,2560,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,2048,0.007088888850477006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,2560,0.03356000118785434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,2560,0.05723377731111315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,1536,0.006274666637182236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,2048,0.028839111328125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,2048,0.049999998675452344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,1536,0.024285333024130926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,1024,0.0058222222659322954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,1536,0.042005333635542125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,1024,0.019296000401178997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,768,0.005241777747869492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,1024,0.03498933381504483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,3584,0.07429955403010051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,768,0.030866665972603693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,512,0.004601777841647466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,512,0.014783999986118741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,512,0.027172444595230952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,256,0.004219555606444676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,256,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,256,0.025285333395004272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,128,0.003946666502290302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,128,0.012655999925401477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,1024,128,0.02164977788925171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,64,0.003539555602603488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,1024,32,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,64,0.012503999802801343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,32,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,1024,768,0.016568000117937725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,65536,0.11664088567097981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,51200,0.09242400195863511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,65536,0.7540853288438586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,51200,0.5909688737657335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,16384,0.031545778115590416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,65536,1.0852897432115343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,16384,0.19620977507697213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,12288,0.023188443647490606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,16384,0.2875280115339491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,12288,0.14564177725050184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,51200,0.8528578016493055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,12288,0.2157555553648207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,10240,0.11890310711330837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,8192,0.016745777593718637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,10240,0.17818933063083223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,8192,0.09311022361119588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,7168,0.014864888456132678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,8192,0.14343378278944227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,7168,0.08233955833646986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,6144,0.013080889152155982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,10240,0.019955555597941082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,7168,0.1277608871459961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,5120,0.011695110963450538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,6144,0.07142933209737141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,6144,0.11209866735670303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,4096,0.009893333746327294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,5120,0.06014044417275322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,5120,0.09670933087666829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,3584,0.008708444734414419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,4096,0.0813244448767768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,4096,0.04841244551870558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,3072,0.008336000144481659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,3584,0.0418853329287635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,3584,0.07368889119890001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,2560,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,3072,0.03742488887574937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,3072,0.06572177675035265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,2048,0.006665777828958299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,2560,0.03219466739230686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,2560,0.05649422274695503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,1536,0.0059057776298787855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,2048,0.049357334772745766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,1536,0.022679999470710754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,1024,0.004936000125275718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,1536,0.04182844360669454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,1024,0.018595554762416415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,768,0.004563555535342959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,1024,0.034767998589409724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,768,0.031151112582948472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,512,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,512,0.01421866648726993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,2048,0.027430221438407898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,256,0.0038480001191298165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,256,0.012798222402731577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,256,0.025404444999165002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,128,0.0035004446076022256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,768,0.016202666693263583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,128,0.012433778080675336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,64,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,128,0.022281777527597215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,768,32,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,64,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,768,512,0.027486221657858953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,768,32,0.011903111305501727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,65536,0.0812515550189548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,51200,0.06560444169574313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,65536,0.7481750912136502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,51200,0.5872595575120714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,16384,0.025054223007626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,65536,1.083804448445638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,16384,0.19499733712938097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,12288,0.020413332515292697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,16384,0.28669510947333443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,12288,0.14434756173027888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,51200,0.8528364499409994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,10240,0.017682666579882305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,12288,0.21539821889665392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,10240,0.11763200494978164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,10240,0.17763288815816244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,8192,0.01567555632856157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,8192,0.09251999855041504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,8192,0.14332444137997097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,7168,0.013199999928474426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,7168,0.0817013316684299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,6144,0.012251555919647217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,6144,0.07061600022845797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,5120,0.01072355525361167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,5120,0.058913780583275684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,6144,0.11259378327263726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,5120,0.09676888916227554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,4096,0.009026666482289633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,7168,0.1275386677847968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,4096,0.04730044470893013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,4096,0.08172711398866442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,3584,0.0417795545525021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,3072,0.007744000189834171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,3584,0.07372799846861097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,3072,0.03605066736539205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,2560,0.007302222152551015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,3072,0.06583999925189547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,2560,0.03187822302182516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,2560,0.05803555250167847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,2048,0.006301333506902059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,3584,0.007979555262459649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,1536,0.005610666755172942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,2048,0.04938933253288269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,1536,0.041456000672446355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,1024,0.004961777892377642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,1024,0.0185208883550432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,1024,0.03474311033884684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,768,0.004263999975389904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,2048,0.027521777484152053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,768,0.015959110524919298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,512,0.003929777691761653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,768,0.030896888838873968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,512,0.014146667387750415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,256,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,256,0.012518222133318583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,256,0.02425066630045573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,128,0.003604444364706675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,1536,0.022822222775883142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,128,0.01218311074707243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,64,0.003289777785539627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,512,0.027836445305082533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,512,128,0.022315555148654517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,512,32,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,65536,0.050817777713139854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,32,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,65536,0.7100702391730415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,51200,0.042018665207756885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,512,64,0.011821333732869891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,65536,1.0836915969848633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,16384,0.016873778568373788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,51200,0.5579386817084419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,16384,0.18643111652798125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,51200,0.8565582169426812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,16384,0.2889920075734456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,12288,0.01406133340464698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,10240,0.013759999639458127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,12288,0.138629330529107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,12288,0.2163155608707004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,10240,0.1127057737774319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,8192,0.012151111331250934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,7168,0.011201777391963534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,10240,0.1784657769733005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,8192,0.08767733309004043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,6144,0.010155555274751451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,8192,0.14476088682810465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,6144,0.11257155736287434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,6144,0.06755733489990234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,5120,0.008966222405433655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,5120,0.09675821993086074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,5120,0.0572355588277181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,4096,0.00776444458299213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,4096,0.08044977982838948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,4096,0.043531556924184166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,3584,0.007287110719415877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,7168,0.12871289253234863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,3072,0.0069875551594628235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,3584,0.03980088896221585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,3584,0.07263466384675768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,2560,0.005979555762476391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,3072,0.034881777233547635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,3072,0.06439199712541369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,2560,0.030143999391131934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,2048,0.005600888695981767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,2560,0.055572443538241915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,1536,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,2048,0.026053332620196875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,2048,0.04893155561553108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,1024,0.004435555388530095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,1536,0.021709332863489788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,1536,0.041026665104760066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,768,0.004072000169091755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,1024,0.01793244481086731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,1024,0.03429155548413595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,7168,0.07875821987787883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,768,0.015480000111791821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,768,0.029840889904234145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,512,0.013517333401574029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,256,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,512,0.02711822258101569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,256,0.012071111136012606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,128,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,256,0.02420622275935279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,128,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,64,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,256,128,0.021047110358874004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,32,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,64,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,65536,0.040123555395338274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,256,512,0.003751111113362842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,256,32,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,51200,0.0331857767370012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,65536,0.7091849115159777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,51200,0.5572657585144043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,16384,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,65536,1.0844649208916557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,16384,0.185442672835456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,12288,0.010067555639478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,16384,0.2866328822241889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,12288,0.1368595494164361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,51200,0.851327101389567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,10240,0.009688888986905416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,12288,0.21464267041948107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,10240,0.1109573311275906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,8192,0.008959111240175035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,10240,0.17737066745758057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,7168,0.008647110727098253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,8192,0.0872124433517456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,8192,0.14322488837771946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,6144,0.009716444545321995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,7168,0.07629688580830891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,7168,0.12663288911183676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,5120,0.008608000146018134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,6144,0.06634133391910128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,6144,0.11102400223414104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,4096,0.007281777759393056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,5120,0.055584887663523354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,5120,0.09563555320103963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,3584,0.0069226668112807805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,4096,0.04487644301520454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,4096,0.0802897744708591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,3072,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,3584,0.03860177927547031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,3584,0.07259288761350843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,2560,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,3072,0.03428622086842855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,3072,0.06463911135991414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,2048,0.00526311124364535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,2560,0.029863112502627905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,2560,0.055732442273033984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,1536,0.00461777796347936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,2048,0.025762667258580525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,2048,0.04869066675504049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,1536,0.02163733376397027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,1536,0.041160888142055936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,1024,0.017860444055663217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,768,0.0038933331767717996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,1024,0.03424711028734843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,768,0.030135999123255413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,768,0.01537333263291253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,512,0.0034977777136696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,512,0.013457777599493662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,512,0.026399110754330952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,256,0.0031644445326593188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,256,0.0236799998415841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,256,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,128,0.0028204443968004654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,128,0.011537777880827585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1536,128,128,0.02128533356719547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,64,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,32,0.0031635556370019913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,64,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,65536,0.03804711169666714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,65536,0.7088515493604871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,51200,0.03338399860594008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,128,1024,0.0042035554846127825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,16384,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,128,32,0.011361777782440186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,12288,0.010055999788973067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,16384,0.1854248841603597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,10240,0.008979555633332994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,51200,0.5567404429117838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,8192,0.008300444318188561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,12288,0.13703021738264295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,7168,0.008001777860853408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,10240,0.11105689075258042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,6144,0.007616000043021307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,8192,0.08693599700927734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,5120,0.00794311116139094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,7168,0.07631555530760023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,6144,0.06583466794755724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,4096,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,3584,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,5120,0.05530844463242424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,3072,0.006093333164850871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,4096,0.043747554222742714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,2560,0.005711110929648082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,3072,0.033732444047927856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,2048,0.004937777916590373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,3584,0.03865422142876519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,1536,0.004534222185611725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,2048,0.025733333494928148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,1024,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,1536,0.0218542218208313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,768,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,768,0.015547555353906421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,512,0.003400000019205941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,2560,0.03003644280963474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,512,0.013887999786270989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,128,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,256,0.012162666353914471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,64,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,1024,0.01789066692193349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,64,0.011493333511882357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,64,32,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,65536,0.03606399893760681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,51200,0.030464887619018555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,65536,0.7088550991482205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,51200,0.557344913482666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,16384,0.015219555960761176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,12288,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,64,32,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,10240,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,12288,0.13707378175523546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,16384,0.18626399834950766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,8192,0.01034044474363327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,10240,0.11072711149851482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,7168,0.009771555662155151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,6144,0.008972444468074376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,7168,0.076273779074351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,8192,0.08746755785412258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,5120,0.008358221915033128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,6144,0.06672444608476427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,4096,0.00795377790927887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,3584,0.0069004446268081665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,5120,0.05458755625618828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,4096,0.042803555727005005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,3072,0.006256000035338932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,3584,0.038303110334608294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,2560,0.005700444595681296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,2048,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,2560,0.030048890246285334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,1536,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,2048,0.025997334056430395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,1024,0.003949333396222857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,1536,0.021607110897699993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,768,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,1024,0.01796800063716041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,768,0.015298666225539314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,512,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,256,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,512,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,256,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,64,0.0027742222365405825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,3072,0.03365155392222934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,128,0.011528000235557556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1536,32,32,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,32,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1536,32,64,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,16384,1.369582176208496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,16384,1.866472880045573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,51200,4.131074693467882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,12288,1.037260479397244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,10240,0.8697351349724664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,12288,1.3292836083306205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,10240,1.0714391072591145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,8192,0.6846008830600314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,8192,0.8446906407674154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,51200,6.23921881781684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,7168,0.5826524628533257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,6144,0.4712550905015733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,7168,0.7597875595092773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,5120,0.401325331793891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,6144,0.6351315710279677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,4096,0.3571840127309163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,5120,0.5247475306193033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,3584,0.2800640000237359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,4096,0.42177155282762313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,3072,0.24248798688252768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,3584,0.4172275596194797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,2560,0.21829156080881754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,3072,0.3176026609208849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,2048,0.16518044471740723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,2560,0.27039554384019643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,1536,0.12790932920244005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,1024,0.09456888834635417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,2048,0.2239182260301378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,1536,0.17688443925645617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,768,0.07371999820073445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,1024,0.13547466860877141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,512,0.05470310979419284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,256,0.042430222034454346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,768,0.1085582243071662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,256,0.08482666810353597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,512,0.09376621908611721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,64,0.03377866744995117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,128,0.08009777466456096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,64,0.08038666513231066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,32,0.03560355636808608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,65536,32,0.08064089218775432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,65536,128,0.03842755489879184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,65536,3.980678982204861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,51200,3.118600845336914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,16384,1.0146133634779189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,65536,6.346719953748916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,16384,1.4664221869574652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,12288,0.7332328690422906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,10240,0.620922671424018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,12288,1.0438177320692275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,8192,0.49655289120144314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,51200,4.9051767985026045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,10240,0.8468444612291125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,7168,0.44863732655843097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,8192,0.67757691277398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,7168,0.5877671241760254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,6144,0.4021475580003526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,5120,0.33841511938307023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,4096,0.2611066765255398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,6144,0.5056959788004557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,5120,0.4204711119333903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,3584,0.22300444708930123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,4096,0.3381982114579942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,3072,0.20010577307807076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,3584,0.29455910788642037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,2560,0.16959288385179308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,2048,0.1314880053202311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,3072,0.25504088401794434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,2560,0.2197128931681315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,1536,0.1026151114039951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,1024,0.07461511426501803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,2048,0.18106667200724283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,768,0.05960444609324137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,1536,0.14330222871568468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,512,0.04376888937420315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,1024,0.10996355613072713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,256,0.033744888173209295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,768,0.08866222037209405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,128,0.030085331863827173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,512,0.07669244209925334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,64,0.02608355548646715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,256,0.06953155332141452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,51200,32,0.027048889133665297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,128,0.06620533598793878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,64,0.06565955612394545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,51200,32,0.06606311268276639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,51200,1.1401626798841689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,65536,1.422025786505805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,16384,0.38172798686557347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,12288,0.2415520085228814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,51200,1.7968603769938152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,16384,0.5384009149339464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,10240,0.20481155978308785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,65536,2.3055288526746964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,8192,0.1642248895433214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,12288,0.4016097651587592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,10240,0.31978045569525826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,7168,0.14108178350660536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,6144,0.12407199541727702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,8192,0.2586231231689453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,5120,0.1065795554055108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,7168,0.22670666376749674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,4096,0.0844808883137173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,6144,0.19660354985131157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,3584,0.07976088921229045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,4096,0.134735107421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,3072,0.06969777743021648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,2560,0.057067553202311196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,3584,0.11955200301276313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,2048,0.04782933327886793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,3072,0.10551289055082534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,1536,0.03821600145763821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,2560,0.09119911326302423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,1024,0.025626666016048853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,2048,0.07693777481714885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,1536,0.06090488698747423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,768,0.02095111045572493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,1024,0.047200000948376127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,512,0.016394666499561734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,256,0.012840888566441007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,768,0.0384133325682746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,128,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,512,0.03219199842876858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,256,0.028170665105183918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,64,0.01035911093155543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,16384,32,0.010399999717871347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,128,0.027099554737408955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,64,0.027124444643656414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,32,0.026908444033728704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,16384,5120,0.16623822848002115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,51200,0.7689733505249023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,65536,0.9921929041544596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,16384,0.2583591143290202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,16384,0.4542844560411241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,51200,1.4529395633273656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,12288,0.19452888435787627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,10240,0.16334756215413412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,65536,1.8441325293646917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,8192,0.12852622403038874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,12288,0.31770488950941295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,8192,0.2092284493976169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,7168,0.11227822303771973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,6144,0.09769066837098862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,7168,0.18446666664547393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,5120,0.08492444621192084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,6144,0.1610364384121365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,4096,0.07074844174914889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,5120,0.13614755206637913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,3584,0.06271821922726102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,4096,0.11219022009107803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,3584,0.09931466976801555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,3072,0.05479200018776787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,2560,0.046358221107059054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,3072,0.08755999803543091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,2048,0.039108445247014366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,2560,0.07583466503355238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,1536,0.031055112679799397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,2048,0.06379733482996623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,1024,0.02375288969940609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,10240,0.26124710506863064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,768,0.020326221982638042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,1536,0.05015822251637777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,1024,0.038754665189319186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,512,0.014754666222466363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,768,0.0325546662012736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,512,0.026956443985303242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,128,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,256,0.02404977712366316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,64,0.00868888861603207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,128,0.023056889573733013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,32,0.009005332986513773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,64,0.023372444841596816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,12288,32,0.023043556345833674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,12288,256,0.011494222614500256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,65536,0.9958711200290256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,51200,0.7620471318562826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,16384,0.25323910183376735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,65536,1.6211990780300563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,12288,0.19199911753336588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,51200,1.271574232313368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,16384,0.3840008841620551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,12288,0.28005867534213597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,10240,0.1586951149834527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,8192,0.12597155570983887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,7168,0.10914222399393718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,10240,0.23151999049716523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,8192,0.18514488803015816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,6144,0.09542844692866008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,5120,0.07964355415768094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,7168,0.16366667217678493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,6144,0.14194399780697292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,4096,0.06522755490409003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,3584,0.057941330803765186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,4096,0.09964178005854289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,5120,0.12097777260674371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,3072,0.05085955394638909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,3584,0.08906844589445327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,2560,0.04296088880962796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,2048,0.03524533245298598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,3072,0.07924977938334148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,2560,0.06781422429614596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,2048,0.05648089117474026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,1024,0.020213334096802604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,1536,0.044954667488733925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,768,0.016501333978441026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,1024,0.035473777188195124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,512,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,768,0.029150221082899306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,256,0.009362666971153682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,512,0.02442933287885454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,128,0.008640000389681922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,256,0.022079111801253423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,128,0.021312889125612047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,64,0.020991999242040846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,1536,0.027375110321574744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,32,0.008629333641793992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,10240,32,0.021351999706692163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,65536,0.7440933121575249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,10240,64,0.00870666652917862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,51200,0.5761440065171983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,16384,0.1877511077457004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,12288,0.1427324480480618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,16384,0.3369857735104031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,10240,0.10545421971215142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,12288,0.24519644843207467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,65536,1.4037866592407227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,8192,0.0880897773636712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,51200,1.101628409491645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,7168,0.07711288664076063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,10240,0.20171199904547799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,6144,0.06545599963929918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,8192,0.1628391080432468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,5120,0.054458667834599815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,7168,0.14333421654171413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,4096,0.045927110645506114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,6144,0.12466311454772949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,3584,0.03969511058595445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,5120,0.10649866527981228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,4096,0.08672800328996445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,3072,0.036885334385765925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,3584,0.07883110973570082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,2560,0.02986577815479702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,2048,0.025024889243973628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,3072,0.06912622186872694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,1536,0.020622221959961783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,2560,0.05888266695870293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,1024,0.014728888869285583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,2048,0.04948444498909844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,1536,0.039784000979529485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,768,0.012621333201726278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,1024,0.031763556930753924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,512,0.00980355590581894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,256,0.00793599999613232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,768,0.026519111461109583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,128,0.007000000112586551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,512,0.022350221872329712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,64,0.0063386667105886675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,128,0.019318222999572754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,256,0.020266667008399963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,8192,32,0.006933333145247565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,64,0.01937599976857503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,8192,32,0.01936533384852939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,65536,0.5823138025071886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,51200,0.43477688895331484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,16384,0.14996000130971274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,65536,1.2921715842352974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,16384,0.31264977984958225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,12288,0.1127671135796441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,10240,0.09477955765194362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,51200,1.0688311258951824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,8192,0.0781537757979499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,12288,0.22897420989142525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,10240,0.1875128878487481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,7168,0.06850222084257337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,8192,0.15113955073886448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,6144,0.0584133333630032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,5120,0.05043199989530775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,7168,0.1334515545103285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,6144,0.11668177445729573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,4096,0.04128888911671109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,3584,0.03694222370783488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,4096,0.08113333251741198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,5120,0.09942133559121026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,3072,0.0325857765144772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,3584,0.07124444511201647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,2560,0.02748888896571265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,2048,0.023357333408461675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,2560,0.05513777666621738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,3072,0.06455644634034899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,1536,0.019313777486483257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,2048,0.04638488756285774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,1024,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,768,0.016209777858522203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,1536,0.03755377729733785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,1024,0.030159999926884968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,512,0.013076444466908773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,768,0.024693333440356787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,256,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,512,0.020968000094095867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,128,0.006656000183688269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,64,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,256,0.01924977699915568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,7168,32,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,128,0.018253333038753934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,64,0.018589332699775696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,7168,32,0.018213333355055917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,65536,0.5102897750006782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,51200,0.3796417713165283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,16384,0.1277608871459961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,16384,0.28835733731587726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,12288,0.09766133626302083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,65536,1.185072898864746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,51200,0.9349528418646919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,10240,0.08387821912765503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,8192,0.06753866540061103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,12288,0.2110382186041938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,7168,0.05959644582536486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,8192,0.1399573352601793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,6144,0.05375022358364529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,7168,0.12356000476413304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,5120,0.043916443983713783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,6144,0.10869955354266697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,5120,0.09220355749130249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,4096,0.03811111052831014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,3584,0.03300177719857957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,4096,0.07526577843560113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,3072,0.028464890188641016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,3584,0.06849777698516846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,10240,0.17372443940904406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,2560,0.02572711143228743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,2048,0.021643555826610986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,3072,0.058154669072892934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,1536,0.01755822201569875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,2560,0.051354666550954185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,2048,0.04352800051371256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,1536,0.034973333279291786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,1024,0.013870221873124441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,768,0.011682666838169098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,1024,0.02835822105407715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,512,0.008692444198661381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,768,0.023363555471102398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,512,0.020059555768966675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,256,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,128,0.006270222365856171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,64,0.005629333357016246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,128,0.017924444542990792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,256,0.01806400054030948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,64,0.017553778158293832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,6144,32,0.005602666487296422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,6144,32,0.017261332935757108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,65536,0.47328890694512266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,51200,0.3668213420444065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,16384,0.1220124430126614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,16384,0.26518577999538845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,65536,1.08753596411811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,12288,0.09329955445395575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,51200,0.858399126264784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,10240,0.08073511388566759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,8192,0.06513422065311007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,12288,0.19414488474527994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,7168,0.05771288606855604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,8192,0.1278995540406969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,6144,0.05005777875582377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,7168,0.11415821976131862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,6144,0.09914666414260864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,5120,0.04232266545295715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,4096,0.0343777769141727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,5120,0.08498666683832805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,3584,0.030407110850016277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,4096,0.07060266865624322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,10240,0.15864266289605036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,3072,0.026387555731667414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,3584,0.062125331825680204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,2560,0.022655111220147874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,2048,0.01918577816751268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,3072,0.05379111236996121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,2560,0.04660444458325704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,1536,0.01570399933391147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,1024,0.011594666375054253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,2048,0.04022044605678982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,1536,0.031932443380355835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,768,0.010036444498433007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,1024,0.025447110335032146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,512,0.007793777518802219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,256,0.006350222147173352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,768,0.021606221795082092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,128,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,512,0.018952889574898612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,64,0.005632888939645555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,256,0.01726222203837501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,128,0.016515556308958266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,64,0.016540444559521146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,5120,32,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,5120,32,0.016252444850073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,65536,0.3504569000667996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,51200,0.2682017750210232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,16384,0.09141955773035686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,16384,0.24115199512905547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,65536,0.9666870964898003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,12288,0.07096266746520996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,51200,0.7641315460205078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,10240,0.058543112542894155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,8192,0.06835999753740099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,10240,0.1444853279325697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,12288,0.17586044470469156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,7168,0.061193777455223926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,8192,0.11697333388858372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,6144,0.05176800158288744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,5120,0.0437466667758094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,7168,0.10410933362113105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,6144,0.09076978100670709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,4096,0.033886220720079206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,5120,0.07733600007163154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,3584,0.030265778303146362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,4096,0.06338666545020209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,3584,0.05564710828993055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,2560,0.02320266597800785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,3072,0.04786044359207153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,2048,0.019073777728610568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,2560,0.04229066769282023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,1536,0.0162471118900511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,2048,0.03640000025431315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,1024,0.010690666735172272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,1536,0.029129776689741347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,768,0.009328888522254096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,1024,0.023237332701683044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,512,0.006994666324721442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,768,0.02030311193731096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,256,0.005650666852792104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,512,0.01756177842617035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,3072,0.0262746661901474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,128,0.0052782222628593445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,64,0.004601777841647466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,256,0.01610933409796821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,128,0.015797333584891427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,64,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,4096,32,0.004884444591071871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,4096,32,0.0151582227812873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,65536,0.3206275569068061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,51200,0.25222934616936576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,16384,0.08735022279951309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,65536,0.9157608879937066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,12288,0.06697244114345975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,16384,0.22986311382717559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,51200,0.7215689023335775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,10240,0.05600800116856893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,8192,0.046132445335388184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,12288,0.16754755708906385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,7168,0.04109333289994134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,10240,0.13736266560024685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,6144,0.03456711106830173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,8192,0.1120106644100613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,7168,0.09895466433631049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,5120,0.02935289012061225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,6144,0.08677955468495686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,4096,0.02392622166209751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,3584,0.021378666162490845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,5120,0.07387110922071669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,3072,0.018963555494944256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,4096,0.060567114088270396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,2560,0.0166302224000295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,3584,0.053448888990614146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,3072,0.04598933458328247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,2048,0.014517333772447375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,1536,0.012068444656001197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,2560,0.040306667486826576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,1024,0.009164444274372524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,2048,0.034837332036760114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,768,0.007975999679830339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,1536,0.02833599845568339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,1024,0.022664889693260193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,512,0.007017778025733099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,256,0.00589155571328269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,768,0.01960266629854838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,128,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,512,0.01720177796151903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,64,0.00452977791428566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,128,0.014905777242448596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,256,0.015641777051819693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3584,32,0.004564444637960858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,64,0.015184889237085978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3584,32,0.014871110518773397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,65536,0.2626613246070014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,51200,0.19212978416019014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,16384,0.06856089168124728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,65536,0.8673404587639703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,51200,0.6825324694315592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,16384,0.21901599566141763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,12288,0.05271910958819919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,10240,0.04317955507172478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,8192,0.034978667894999184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,12288,0.15930933422512478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,10240,0.13090399901072183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,7168,0.0326204432381524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,6144,0.029456890291637842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,8192,0.10620355606079102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,7168,0.09433777464760675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,5120,0.023949333363109167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,6144,0.08218310938941108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,4096,0.020403555697864957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,3584,0.01867911054028405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,5120,0.07015999820497301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,4096,0.0564862224790785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,3072,0.016515556308958266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,2560,0.014887111054526435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,3072,0.0439591109752655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,3584,0.05084977878464592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,2048,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,2560,0.03850400116708543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,1536,0.010078222387366826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,1024,0.008292444050312042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,2048,0.03305155701107449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,1536,0.02693155573474036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,1024,0.021672000487645466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,768,0.007356444166766272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,512,0.006284444282452266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,256,0.0052382221652401825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,768,0.018908444378111098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,128,0.004574222283230888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,512,0.01687111126052009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,256,0.015187554889255099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,64,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,3072,32,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,128,0.01480888823668162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,64,0.014469333820872836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,3072,32,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,65536,0.2503306600782606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,51200,0.19640444384680855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,16384,0.06705155637529162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,65536,0.8191671371459961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,51200,0.6452755398220485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,12288,0.05144266618622673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,16384,0.20778400368160674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,10240,0.04418311185306973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,12288,0.15154843860202366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,8192,0.03425600131352743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,10240,0.12433421611785889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,7168,0.029437333345413208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,8192,0.10066044330596924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,6144,0.02502577834659153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,7168,0.08929155270258586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,5120,0.021689777572949726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,6144,0.07835021946165296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,4096,0.017887110511461895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,5120,0.06624266836378309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,3584,0.01645955608950721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,4096,0.054098665714263916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,3072,0.01444533301724328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,3584,0.04734844300482008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,2560,0.012917333179050021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,3072,0.041834665669335254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,2560,0.036481777826944985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,2048,0.011027555498811932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,1536,0.008985777695973715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,1024,0.007310222420427535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,1536,0.025431111454963684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,2048,0.03174044357405768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,768,0.006568000134494569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,1024,0.020617778102556866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,512,0.0058133333093590206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,256,0.004889777965015835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,768,0.01796533332930671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,512,0.015832000308566622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,128,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,256,0.01442577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,64,0.004228444563017951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2560,32,0.004323555363549126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,128,0.014150222142537435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,64,0.014112000664075216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2560,32,0.013753777576817406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,65536,0.19417866071065268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,51200,0.15065244833628336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,16384,0.05199288990762499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,16384,0.1970613267686632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,12288,0.03973955578274197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,51200,0.606677320268419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,65536,0.775418652428521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,10240,0.03345688846376207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,8192,0.027944889333513048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,12288,0.14377066824171278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,10240,0.11798400349087185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,7168,0.023772444989946153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,8192,0.09538933303621079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,6144,0.020777778493033517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,5120,0.01816444430086348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,7168,0.0846595565478007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,6144,0.07356355587641399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,4096,0.015013333823945789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,3584,0.01347644461525811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,5120,0.062257779969109424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,3072,0.012673777838548025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,4096,0.04948711064126757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,3584,0.044728888405693897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,2560,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,2048,0.009313777916961247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,2560,0.03442755672666762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,2048,0.029119110769695703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,1536,0.007713778151406183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,1024,0.00646311127477222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,1536,0.024139554964171514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,768,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,1024,0.02014044423898061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,512,0.005195555587609609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,768,0.01718222267097897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,256,0.004491555607981152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,512,0.015509333875444202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,128,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,256,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,64,0.003842666745185852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,128,0.013760888742076026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,2048,32,0.003839111162556542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,3072,0.039332442813449435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,64,0.013400889105266996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,2048,32,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,65536,0.19545688894059923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,51200,0.12927199734581843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,16384,0.044454223579830594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,16384,0.18612444400787354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,51200,0.5698906580607096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,65536,0.7259759902954102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,12288,0.032997333341174655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,10240,0.036678221490648054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,8192,0.027161777019500732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,12288,0.13607554965549046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,7168,0.026017778449588354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,8192,0.09010844760470921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,6144,0.020989333589871723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,7168,0.0799235569106208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,5120,0.01888177792231242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,6144,0.06918222374386258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,4096,0.016565332810084026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,5120,0.058463109864128955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,3584,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,4096,0.046903110212749906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,3584,0.04238933324813843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,3072,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,10240,0.11140622033013238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,2560,0.010223111344708337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,2048,0.008372444245550368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,2560,0.03189155459403992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,1536,0.007334222396214803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,2048,0.028845333390765723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,1536,0.023060444328520033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,1024,0.006637333167923822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,768,0.006269333263238271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,1024,0.019311999281247456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,512,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,768,0.01695644524362352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,256,0.004416888786686791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,512,0.015174221661355762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,128,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,3072,0.036997334824668035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,64,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,256,0.013817778064144982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1536,32,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,64,0.012849777936935425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,32,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,65536,0.14315822389390734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,51200,0.11133244302537705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,51200,0.5338995721605089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,65536,0.6813439793056912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,16384,0.038479109605153404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1536,128,0.013155555559529198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,12288,0.029095109966066148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,10240,0.024650666448805068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,16384,0.17553244696723092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,10240,0.10476266675525242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,8192,0.08421866761313544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,7168,0.018171555466122098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,7168,0.07450755437215169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,6144,0.016167999969588388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,6144,0.06438044706980388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,12288,0.1273555490705702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,5120,0.013832888669437833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,4096,0.011710222396585675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,8192,0.020647999313142564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,3584,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,5120,0.05400177836418152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,4096,0.0435751113626692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,3072,0.009693333672152625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,2560,0.008729777402347988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,3072,0.0344453321562873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,3584,0.03983644313282437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,2048,0.007672000262472365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,2560,0.030489779180950586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,1536,0.00673333348499404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,1024,0.0053084443012873335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,1536,0.02200355629126231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,2048,0.02686133318477207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,768,0.00489155575633049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,1024,0.018215111560291715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,512,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,256,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,512,0.014189332723617554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,128,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,768,0.015992000699043274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,256,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,64,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,1024,32,0.0033564445459180405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,128,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,64,0.012482666307025485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,65536,0.1147644387351142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,51200,0.09200355741712783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,1024,32,0.012375999655988483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,16384,0.031515555249320135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,65536,0.645219537946913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,16384,0.16647466023763022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,12288,0.024077332682079736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,51200,0.504351986779107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,10240,0.021002666817771062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,8192,0.017648888958825003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,10240,0.09900177849663629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,12288,0.121796449025472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,7168,0.016093333562215168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,8192,0.07989866866005792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,6144,0.01423377792040507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,5120,0.012647111382749347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,7168,0.07091377841101752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,6144,0.06117510795593262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,4096,0.010455999937322406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,5120,0.05056266652213203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,3584,0.00944622192117903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,3072,0.008685333033402761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,4096,0.041746666034062706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,2560,0.007602666815121968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,3072,0.03269244564904107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,2560,0.028519110547171697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,2048,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,1536,0.006256000035338932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,2048,0.025174222058720056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,1024,0.005245333330498801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,1536,0.02108977735042572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,768,0.00490399988161193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,1024,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,768,0.015503111812803479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,512,0.004535111288229625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,3584,0.037159111764695905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,256,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,512,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,128,0.003539555602603488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,256,0.012532444463835822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,64,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,128,0.012410666379663678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,768,32,0.0033280000918441345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,64,0.012126222252845764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,65536,0.08224444256888495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,768,32,0.012087999946541257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,51200,0.06627377536561754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,16384,0.0258942229880227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,16384,0.16549421681298151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,12288,0.021202666891945735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,51200,0.5021422174241807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,65536,0.6391644477844238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,10240,0.016354666815863717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,8192,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,12288,0.12046221892038982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,10240,0.09797777732213338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,7168,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,6144,0.011030221978823343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,8192,0.07920800315009223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,5120,0.00962222201956643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,7168,0.07023111316892836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,4096,0.01001422190003925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,6144,0.06004889143837822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,5120,0.050556444459491305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,4096,0.04052711195415921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,3072,0.008056888977686564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,3584,0.036496887604395546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,2560,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,3072,0.0321048895517985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,2048,0.006583111153708563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,2560,0.028315554062525432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,1536,0.005596444424655702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,2048,0.02499377727508545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,1024,0.004852444347408083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,1536,0.020891555481486853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,768,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,1024,0.017616889543003507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,3584,0.008639111287064022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,512,0.003900444342030419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,768,0.015454222758611044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,256,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,128,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,512,0.01348000019788742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,256,0.012492444780137805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,64,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,512,32,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,128,0.011882666912343768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,65536,0.049991998407575816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,64,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,512,32,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,51200,0.041061331828435264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,16384,0.015229332778188916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,16384,0.15644444359673396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,12288,0.014137778017255994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,51200,0.47203021579318577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,65536,0.6007093323601617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,10240,0.013043555948469373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,8192,0.012095111111799875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,12288,0.11241243945227729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,10240,0.09184622102313572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,7168,0.010766222245163388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,6144,0.009716444545321995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,8192,0.07384088966581556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,5120,0.00868000007337994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,7168,0.06541688574684991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,4096,0.007680888805124495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,6144,0.05641333262125651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,3584,0.007037333316273159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,5120,0.04638311266899109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,3072,0.006626666833957036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,4096,0.038110223081376814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,2560,0.005991999887757831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,3584,0.033979554971059166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,2048,0.0052151112920708126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,3072,0.0302746660179562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,1536,0.004938666605287128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,2560,0.02645333276854621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,1024,0.004263999975389904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,2048,0.023970666858885024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,1536,0.01995377739270528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,768,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,512,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,1024,0.016557332542207506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,256,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,768,0.01479200025399526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,128,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,512,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,64,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,256,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,256,32,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,128,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,64,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,65536,0.03447377681732178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,256,32,0.01130488928821352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,51200,0.03239288926124573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,65536,0.6004453235202366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,51200,0.4717404577467177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,16384,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,12288,0.010693333215183683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,16384,0.15610755814446345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,12288,0.11253155602349176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,10240,0.009707555174827576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,8192,0.009145778086450364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,10240,0.09199111329184638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,7168,0.009321778184837764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,8192,0.07383822070227729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,6144,0.008829333715968663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,7168,0.06517510943942599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,5120,0.008692444198661381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,6144,0.05575111177232531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,4096,0.007576000359323289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,5120,0.04626577761438158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,3584,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,4096,0.038022223446104265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,3072,0.00657155571712388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,3584,0.03330488999684652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,2560,0.005904888941182031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,3072,0.03017510970433553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,2048,0.0052622221410274506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,2560,0.026825777358478967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,1536,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,2048,0.023648000425762598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,1024,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,1536,0.01960355540116628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,768,0.003818666769398583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,1024,0.016574222180578444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,512,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,768,0.014462222655614218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,256,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,512,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,128,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,256,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,64,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,128,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,128,32,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,64,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,65536,0.03259733319282532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,128,32,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,51200,0.02762577765517765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,51200,0.4717671076456706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,65536,0.600147565205892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,12288,0.009950222240553962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,16384,0.011895111037625207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,10240,0.009394666386975182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,16384,0.15531822045644125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,8192,0.00831288927131229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,12288,0.11283377806345622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,7168,0.007974222302436829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,10240,0.09188889132605658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,8192,0.07354666789372762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,6144,0.007686222592989604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,7168,0.06473600202136569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,5120,0.008339555727110969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,4096,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,6144,0.05599911345375908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,5120,0.04638933473163181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,3584,0.006649777707126405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,3072,0.006329777754015393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,4096,0.03736889031198289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,3584,0.03326311045222812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,3072,0.02990488873587714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,2560,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,2048,0.004982222285535601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,1536,0.00462400002612008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,2560,0.02658133374320136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,1024,0.003932444585694207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,2048,0.023761777414215937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,1536,0.019926221834288705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,768,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,512,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,1024,0.016551110479566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,256,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,768,0.014492443866199918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,512,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,256,0.01180622229973475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,64,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,128,0.011438222395049201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,64,32,0.0029502221279674103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,64,0.011141333315107556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,65536,0.031186666753556993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,51200,0.025935999221271936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,64,32,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,16384,0.015012444721327888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,16384,0.1553537713156806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,12288,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,65536,0.6002417670355903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,51200,0.4718880123562283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,10240,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,8192,0.010039110978444418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,12288,0.11256533198886448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,7168,0.010727110836240979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,8192,0.07344533337487115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,7168,0.06493333313200209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,6144,0.009678222239017487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,6144,0.055177778005599976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,5120,0.00869333330127928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,5120,0.04613777663972643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,4096,0.007630222373538547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,10240,0.09199021922217475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,4096,0.03772622346878052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,3584,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,3072,0.006628444625271692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,3584,0.03330311179161072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,2560,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,3072,0.030260443687438965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,2048,0.005403555515739653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,1536,0.004280000097221798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,2560,0.02664711078008016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,1024,0.0038444445365005066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,2048,0.023744889431529578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,1536,0.019864888654814828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,768,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,512,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,1024,0.016495111915800307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,256,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,768,0.01479288935661316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,512,0.013428444663683573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,256,0.01181688904762268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,64,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1281,32,32,0.0028115556471877625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,128,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,64,0.011113777756690979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1281,32,32,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,51200,3.9377644856770835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,16384,1.2552959654066298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,51200,5.896339416503906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,16384,1.7708027097913954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,12288,0.9430916044447156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,12288,1.2457271152072482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,16384,1.327090687221951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,51200,4.206481085883246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,10240,0.8621733453538684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,10240,1.0255200068155925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,12288,0.9969431559244791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,8192,0.6903066635131836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,8192,0.8086160024007162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,7168,0.6014915572272407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,10240,0.8361484209696451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,7168,0.6993458006117078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,8192,0.679382218254937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,6144,0.5526435640123155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,6144,0.5996630986531576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,5120,0.40586400032043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,7168,0.59861511654324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,4096,0.33025511105855304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,5120,0.49907557169596356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,4096,0.42714844809638125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,5120,0.4427911175621881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,3584,0.2963342136806912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,4096,0.36506933636135525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,3584,0.35269689559936523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,3072,0.24217510223388672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,6144,0.5213955773247613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,3584,0.3242986732059055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,2560,0.2152257760365804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,3072,0.30622487597995335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,2048,0.16704711649152967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,3072,0.28676533699035645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,2560,0.26171554459465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,1536,0.12797510623931885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,2560,0.24607290161980522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,2048,0.2161804437637329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,2048,0.20856711599561903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,1024,0.08674399720297919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,1536,0.17060711648729113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,1536,0.16850222481621635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,768,0.06883466906017728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,1024,0.13278667132059732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,768,0.10510577758153279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,512,0.055762668450673424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,768,0.11150044865078396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,256,0.04039377636379666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,512,0.09077599975797866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,512,0.09336177508036296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,128,0.03333066569434272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,256,0.07574133078257243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,128,0.07759644587834676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,65536,128,0.06196977694829305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,64,0.03032711148262024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,65536,32,0.03187822302182516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,1024,0.13104711638556585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,64,0.0772248903910319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,32,0.07798044549094306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,65536,256,0.08199910985098945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,51200,3.086714638604058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,65536,3.899108462863498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,65536,5.924724578857422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,16384,0.9834435780843099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,16384,1.3768942091200087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,51200,4.731259663899739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,65536,4.385514577229817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,12288,0.7815555466545953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,16384,1.0941599739922416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,51200,3.431211471557617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,12288,0.9944399727715386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,10240,0.6782062318589953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,10240,0.8084550963507758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,12288,0.8258577982584635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,8192,0.5301742023891872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,8192,0.6478346718682183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,10240,0.6902222103542752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,7168,0.4623689121670193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,7168,0.5587777561611599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,8192,0.5595191319783529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,6144,0.3988106780582004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,7168,0.4952079984876845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,6144,0.4824479950798883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,5120,0.3321271207597521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,5120,0.4016142156389024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,6144,0.430656009250217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,4096,0.2598862118191189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,5120,0.36672088834974503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,4096,0.3217359913720025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,3584,0.23276178042093912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,4096,0.3024657832251655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,3584,0.28181778060065377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,3072,0.20665421750810412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,3072,0.24581421746148002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,3584,0.2689013216230604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,2560,0.16491377353668213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,3072,0.2372035450405545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,2560,0.21101600593990752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,2048,0.14230222172207302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,2560,0.20523644818200004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,2048,0.17484266228146025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,1536,0.10609422127405803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,2048,0.17432355880737305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,1536,0.1383662223815918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,1536,0.14108888308207193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,1024,0.10666222042507595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,1024,0.11185599697960748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,768,0.05911644299825033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,512,0.045287112394968666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,768,0.09413688712649876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,768,0.08583288722568089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,512,0.07425600290298462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,256,0.032481776343451604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,512,0.07904533545176189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,128,0.027740443746248882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,256,0.0671022203233507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,256,0.06501866711510552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,64,0.024159111910396155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,128,0.06335288948482938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,51200,128,0.053835554255379565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,1024,0.07783822218577068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,51200,32,0.024884444144037034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,64,0.0637964473830329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,51200,32,0.06366399923960368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,51200,0.9764337539672852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,65536,1.3133243984646268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,65536,2.2052408854166665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,16384,0.3173244529300266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,51200,1.7193351321750219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,16384,0.5229182243347168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,12288,0.24040799670749238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,65536,1.966917249891493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,16384,0.5098640124003092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,12288,0.37543556425306535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,10240,0.2164613273408678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,51200,1.5464195675320094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,10240,0.30791378021240234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,12288,0.3800533347659641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,8192,0.1937075588438246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,7168,0.16478578249613443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,10240,0.32003021240234375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,8192,0.24628355767991808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,8192,0.25992266337076825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,6144,0.1406373315387302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,7168,0.2212808926900228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,5120,0.12000355455610488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,6144,0.1895688904656304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,7168,0.23151644070943198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,6144,0.2026231156455146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,4096,0.09058311250474717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,5120,0.16038755575815836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,5120,0.17356267240312365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,4096,0.13069955507914224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,3584,0.08621599939134385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,4096,0.14450489150153265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,3072,0.07405155234866671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,3584,0.1159671147664388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,3584,0.13002667162153456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,2560,0.05999466445710924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,3072,0.10242755545510186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,2048,0.051141334904564746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,3072,0.11633066336313884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,2560,0.08848977751202053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,2560,0.10150933265686035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,1536,0.039938665098614164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,2048,0.07449599769380358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,1536,0.05923110908932156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,1536,0.0728035569190979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,2048,0.08849600288603042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,1024,0.02796799937884013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,768,0.02253155575858222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,1024,0.04554577668507894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,1024,0.06034488810433281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,768,0.037055111593670316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,512,0.01761511133776771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,768,0.05212622218661838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,256,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,512,0.031389332479900785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,512,0.04562755425771078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,128,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,256,0.02760533326201969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,256,0.03742311067051358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,64,0.009670221971140968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,16384,128,0.03117777903874715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,16384,32,0.010035555395815108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,128,0.026766222384240892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,64,0.02642400066057841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,16384,32,0.02606311109330919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,65536,1.0837412940131295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,65536,1.7610639995998805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,51200,0.8329831229315864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,65536,1.7151075998942058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,51200,1.3828577465481227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,16384,0.2767057683732775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,16384,0.41944532924228245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,12288,0.2098062170876397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,16384,0.44387200143602157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,51200,1.3425351248847113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,12288,0.30668889151679146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,10240,0.17071999443901908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,12288,0.33104001151190865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,10240,0.25169867939419216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,8192,0.1368746625052558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,10240,0.28349243270026314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,8192,0.20250844955444336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,7168,0.11876444021860759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,8192,0.22684534390767416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,6144,0.10565688874986436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,7168,0.1781991057925754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,7168,0.20167377259996203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,5120,0.08580355511771308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,6144,0.15499999788072374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,6144,0.17593599690331352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,4096,0.07052444749408297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,5120,0.13244356049431696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,5120,0.150964445537991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,3584,0.06387466854519315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,4096,0.1089715560277303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,4096,0.12682578298780653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,3584,0.0969395571284824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,3584,0.11409689320458306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,3072,0.0529022216796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,2560,0.0445128877957662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,3072,0.08589066399468316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,3072,0.10214310884475708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,2048,0.0373964442147149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,2560,0.07478400071461995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,2560,0.08984266387091742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,1536,0.028532445430755615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,2048,0.06176622046364678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,2048,0.07862044705284967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,1536,0.04909955461819967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,1536,0.06499377886454265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,1024,0.020282665888468426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,768,0.016598222984208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,1024,0.03769066598680284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,1024,0.0540559987227122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,512,0.01475288967291514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,768,0.031480001078711614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,768,0.04680977927313911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,256,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,512,0.02679555614789327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,512,0.0409786668088701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,256,0.0236124445994695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,256,0.033194667763180204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,128,0.009020444419648912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,64,0.011233777635627322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,12288,128,0.029920889271630183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,128,0.022661333282788593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,12288,32,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,64,0.02314044369591607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,12288,32,0.022348443667093914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,65536,0.8280311160617404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,51200,0.6529368824428982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,65536,1.5659768846299913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,16384,0.21502578258514404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,51200,1.2208293279012044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,16384,0.3750782277848985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,12288,0.16302400165134007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,65536,1.5838498009575739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,16384,0.4119119909074571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,51200,1.2453163994683158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,10240,0.1490079959233602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,12288,0.2724977864159478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,12288,0.30767732196384007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,8192,0.11724000506930882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,10240,0.22385954856872559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,10240,0.2573546568552653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,7168,0.10902488893932766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,8192,0.17998488744099936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,8192,0.2096479998694526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,6144,0.09291022353702122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,7168,0.15856533580356175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,7168,0.1868364413579305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,6144,0.1379271083407932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,5120,0.07569866710238986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,6144,0.1632631089952257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,4096,0.0618035528394911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,5120,0.11652799447377522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,5120,0.14174044132232666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,4096,0.09751288758383857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,4096,0.11754133966233994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,3584,0.05455999904208713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,3072,0.04668000009324816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,3584,0.08704533179601033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,3072,0.07737955782148573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,3072,0.09560266468260024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,2560,0.03983911209636264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,2560,0.06604977448781331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,2560,0.08449777629640366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,2048,0.03283199999067519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,2048,0.07292533583111234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,2048,0.05556977457470364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,1536,0.02568088968594869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,3584,0.10642311308119033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,1536,0.06117600202560425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,1024,0.019766221443812054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,1024,0.03497777713669671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,768,0.015955555770132277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,768,0.02852977646721734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,768,0.044248888889948525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,512,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,512,0.024432889289326135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,1536,0.04418933391571045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,256,0.010035555395815108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,512,0.03746044304635789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,1024,0.05045688814587063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,128,0.00871733327706655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,256,0.03221333358022902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,256,0.021894221504529316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,128,0.0207022229830424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,64,0.008392000363932716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,10240,32,0.00868444475862715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,10240,128,0.028538667493396338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,64,0.020658666888872784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,10240,32,0.02075466679202186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,65536,0.7384506861368815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,51200,0.573304017384847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,65536,1.3481839497884114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,51200,1.059366226196289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,65536,1.4546257654825847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,16384,0.19602666960822213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,16384,0.326986657248603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,51200,1.1430435180664062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,16384,0.3795493443806966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,12288,0.14967555469936794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,10240,0.11380799611409505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,12288,0.23807289865281847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,12288,0.281584898630778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,8192,0.09390044212341309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,10240,0.19639733102586535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,10240,0.23675200674268934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,7168,0.08116977744632296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,8192,0.15790755218929714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,8192,0.19314755333794487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,7168,0.13964533805847168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,6144,0.06944355699751112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,7168,0.17115555869208443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,5120,0.058429333898756236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,6144,0.12114755312601726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,4096,0.045873777733908765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,6144,0.15111645062764487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,5120,0.10307111342748006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,5120,0.13034221861097547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,3584,0.04298133320278592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,4096,0.08410400152206421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,4096,0.1093439989619785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,3072,0.037808888488345675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,3584,0.07657155725691053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,2560,0.030244443151685927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,3072,0.08835377958085801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,2560,0.05737244420581394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,3072,0.06785333156585693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,2048,0.025166221790843542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,2560,0.0778213342030843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,1536,0.020779555042584736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,2048,0.06785866949293348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,2048,0.04870666729079353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,1536,0.03886311252911886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,1024,0.015170666906568738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,1536,0.05706933471891615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,768,0.012684444586435953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,1024,0.030249777767393324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,1024,0.04818577898873223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,512,0.009914666414260864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,768,0.025467554728190105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,768,0.03983555568589105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,3584,0.09833511379030015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,256,0.007631111476156447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,512,0.022050667140218947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,512,0.03495111068089803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,128,0.006980444822046492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,256,0.01960177719593048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,256,0.03094044327735901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,64,0.006575999988449945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,128,0.01900533338387807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,8192,128,0.02719555629624261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,8192,32,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,32,0.018810666269726224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,8192,64,0.019359111785888672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,65536,0.5444266531202528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,51200,0.42710222138298887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,65536,1.2542497846815321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,51200,0.9824399948120117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,65536,1.326364411248101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,16384,0.14550666014353433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,16384,0.3047564559512668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,16384,0.3470124403635661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,12288,0.11443289120992024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,51200,1.041776869032118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,10240,0.10451110866334702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,12288,0.22219822141859266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,12288,0.2583368884192573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,8192,0.081968002849155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,10240,0.18247821595933703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,10240,0.21641155083974203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,7168,0.06908533308241102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,8192,0.1761324405670166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,8192,0.1466053326924642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,7168,0.13020177682240805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,6144,0.06162933508555094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,7168,0.15735822253757054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,5120,0.05229333374235365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,6144,0.11383110947079128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,6144,0.13929243882497153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,4096,0.04196444484922621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,5120,0.11985777484046088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,4096,0.07950399981604682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,4096,0.10043111112382676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,3584,0.03756177756521437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,3584,0.06879378027386136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,3584,0.09144622087478638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,3072,0.03310133351220025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,3072,0.08153333266576131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,2560,0.028318223026063707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,3072,0.0635991096496582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,5120,0.09618488947550456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,2048,0.023710222707854375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,2560,0.054135110643174916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,2560,0.07292533583111234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,2048,0.045273777511384755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,1536,0.019267555740144517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,2048,0.06317155228720771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,1024,0.014858666393491956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,1536,0.03667733404371474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,1536,0.0539226664437188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,1024,0.029162665208180744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,768,0.014634667171372307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,1024,0.04520977867974175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,512,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,768,0.023919999599456787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,768,0.03927288783921136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,256,0.0069671107663048645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,512,0.02102577851878272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,512,0.03322666552331712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,128,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,256,0.02982577681541443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,128,0.01793688866827223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,256,0.018947554959191214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,64,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,7168,128,0.026745777991082933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,7168,32,0.006638222270541721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,64,0.018287110659811232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,7168,32,0.01752355529202355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,65536,0.5197431246439616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,51200,0.40803199344211155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,65536,1.14591736263699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,16384,0.13708710670471191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,51200,0.9027840296427408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,16384,0.28250400225321454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,65536,1.325287077162001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,12288,0.10562044382095337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,12288,0.20559110906389025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,16384,0.3463875452677409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,10240,0.09686222341325547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,12288,0.25816088252597386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,10240,0.1687306695514255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,8192,0.0772542224989997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,10240,0.21570400396982828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,7168,0.06906400124231975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,8192,0.1359599961174859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,8192,0.17644711335500082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,6144,0.059234665499793164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,7168,0.12053867181142171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,7168,0.15643644332885742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,5120,0.05029511120584276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,6144,0.10569066471523708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,6144,0.13803288671705458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,5120,0.11932355827755398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,51200,1.0421555836995442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,4096,0.04055111275778876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,5120,0.09005866448084514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,3584,0.03641244437959459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,4096,0.07469866673151652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,4096,0.0998399986161126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,3072,0.03199822372860379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,3584,0.06650577651129828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,3072,0.056074665652381055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,2560,0.027318222655190363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,3072,0.08164444234636095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,2560,0.04964977833959791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,2048,0.02291822267903222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,2560,0.07244800196753608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,2048,0.04211555586920845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,1536,0.01856355534659492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,2048,0.06360089116626315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,1536,0.03419111172358195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,1536,0.05340622199906243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,1024,0.014499555031458536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,3584,0.09133599864112006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,768,0.011256888508796692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,1024,0.04459555612670051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,1024,0.027338667048348322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,768,0.02239466706911723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,512,0.009445333646403419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,768,0.0374035570356581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,256,0.007332444190979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,512,0.01926933394538032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,512,0.033238222201665245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,256,0.028925332758161757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,256,0.017903111047214933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,128,0.006416000011894438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,64,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,6144,128,0.025803556044896443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,128,0.017265778448846605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,6144,32,0.005610666755172942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,64,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,6144,32,0.01717333330048455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,65536,0.4712248908148871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,51200,0.368376890818278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,65536,1.046555519104004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,51200,0.8214462068345812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,65536,1.1968213187323677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,16384,0.12350489033593072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,16384,0.2600231170654297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,12288,0.09470755524105495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,16384,0.31400712331136066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,51200,0.9423021740383573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,10240,0.07592888673146565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,12288,0.18979644775390625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,12288,0.23291201061672637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,8192,0.05495377712779575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,10240,0.19562933180067274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,10240,0.15571288267771402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,7168,0.05231377813551161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,8192,0.12582043806711832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,8192,0.15994844171735975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,6144,0.04497244291835361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,7168,0.14265067047542995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,7168,0.11181422074635823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,6144,0.0978613363371955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,5120,0.03683822353680929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,6144,0.12570844756232366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,4096,0.030750221676296655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,5120,0.08380266692903306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,5120,0.1094471083747016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,4096,0.06910755236943562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,4096,0.09140266974767049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,3584,0.06110666857825386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,3584,0.08328266938527425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,3072,0.023875556058353845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,3072,0.0522017776966095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,3072,0.07496800025304158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,2560,0.021363554729355708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,2560,0.045257776975631714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,2048,0.0170133329100079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,3584,0.027512888113657635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,2048,0.03910488883654276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,1536,0.013811555173661975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,2048,0.05900444587071737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,1536,0.03180888957447476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,1024,0.010553778045707278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,1536,0.04979644550217522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,1024,0.025666667355431452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,768,0.00870400004916721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,1024,0.04243466589185926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,2560,0.06626755661434598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,768,0.02130311065249973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,768,0.03541333476702372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,512,0.018586667047606576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,256,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,512,0.03129511078198751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,256,0.01687111126052009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,256,0.02916355596648322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,128,0.005636444522274866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,64,0.0052764444715446895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,5120,128,0.025424000289705064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,128,0.016618667377365958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,32,0.005622222191757626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,64,0.01629866659641266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,5120,512,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,5120,32,0.016169778174824186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,65536,0.363123549355401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,51200,0.28565777672661674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,65536,0.9459066390991211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,51200,0.7458693186442057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,16384,0.10010400083329941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,65536,1.1984995736016166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,16384,0.23668355411953398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,12288,0.07543377743826972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,51200,0.9415351019965278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,16384,0.314431111017863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,10240,0.06117244561513265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,12288,0.17317067252265084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,12288,0.23276087972852919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,8192,0.04673155479960971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,10240,0.1429493294821845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,10240,0.19464443789588082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,7168,0.04262488749292162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,8192,0.11558311515384251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,8192,0.15973688496483698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,7168,0.10287377569410537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,6144,0.03748622205522325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,7168,0.142992893854777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,5120,0.032602667808532715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,6144,0.09029777844746907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,6144,0.12638933128780788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,4096,0.02570666703912947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,5120,0.07665333482954237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,5120,0.10876978105968899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,3584,0.023027555810080633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,4096,0.09106310870912339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,4096,0.0636319981680976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,3584,0.05515911181767782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,3584,0.08260177903705172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,3072,0.02029244436158074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,2560,0.017469333277808297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,3072,0.04780800143877665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,3072,0.07427910963694255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,2048,0.015480889214409722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,2560,0.04217155443297493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,2560,0.06607911321851942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,1536,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,2048,0.036228444841172956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,2048,0.05856088797251383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,1024,0.00923111124171151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,1536,0.029412445094850328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,1024,0.03944355580541823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,768,0.008018666671382057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,1024,0.023331556055280898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,768,0.019983111156357657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,768,0.035346666971842446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,512,0.006959111326270633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,512,0.03056710958480835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,256,0.005919111271699269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,256,0.015965332587560017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,1536,0.04854222138722738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,128,0.005246222019195557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,256,0.027782221635182697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,128,0.015255111787054272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,64,0.0052737775776121355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,4096,128,0.0240328891409768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,4096,32,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,64,0.015170666906568738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,32,0.015174221661355762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,4096,512,0.017588444881969027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,65536,0.2987911171383328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,51200,0.22724355591668022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,65536,0.9024782180786133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,51200,0.715520011054145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,16384,0.08284711175494723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,65536,1.0710453457302518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,16384,0.22747733857896593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,12288,0.06068978044721815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,51200,0.8404311074150933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,16384,0.28290221426222056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,10240,0.05131555596987406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,12288,0.16523910893334284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,8192,0.041794667641321816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,12288,0.2093235519197252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,10240,0.13665066825018987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,10240,0.17507644494374594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,7168,0.03688800003793504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,8192,0.14373155434926352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,7168,0.09860177834828694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,6144,0.033117334047953285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,7168,0.12822666433122423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,5120,0.027577777703603108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,6144,0.0858702196015252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,5120,0.07363110780715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,5120,0.09856444597244263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,4096,0.05965155363082886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,8192,0.11026933458116318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,3584,0.020695999264717102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,4096,0.08306311236487494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,3584,0.052228444152408175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,3584,0.07540888918770684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,3072,0.018410666121376883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,6144,0.11374844445122613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,3072,0.04594044552909004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,3072,0.06806577576531304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,2560,0.016397333807415433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,2560,0.06076444519890679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,2560,0.04018133216434055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,2048,0.013791999883121915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,2048,0.03428355521625943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,2048,0.0536480016178555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,1536,0.011537777880827585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,1536,0.04388533367051018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,1024,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,1024,0.022608000371191237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,4096,0.02295111119747162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,768,0.007685333490371704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,1024,0.036991112762027316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,768,0.01941511034965515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,768,0.032941334777408175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,512,0.00655644428398874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,512,0.016546666622161865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,256,0.005299555758635203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,512,0.029863999949561224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,1536,0.02776977750990126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,128,0.004919111314747068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,256,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,256,0.027215111586782668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,64,0.004301333179076512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,128,0.014879110786649915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3584,128,0.023073777556419373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3584,32,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,64,0.014852444330851236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3584,32,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,65536,0.2775457700093587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,51200,0.22013244363996717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,65536,0.8599342240227593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,51200,0.6740319993760852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,16384,0.07399377557966444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,65536,1.0687902238633897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,16384,0.21522755093044707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,12288,0.05687910980648465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,16384,0.28201866149902344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,51200,0.841166231367323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,12288,0.15785245100657144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,12288,0.20865955617692736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,10240,0.04923199945025974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,8192,0.04012622104750739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,10240,0.12990310457017687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,10240,0.17486400074428984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,7168,0.035750223530663386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,8192,0.1432160006629096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,8192,0.10550844669342041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,6144,0.030927111705144245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,7168,0.09382310840818618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,7168,0.12844267156389025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,5120,0.02678044471475813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,6144,0.1134106715520223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,6144,0.08229421907001071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,4096,0.02232088810867733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,5120,0.07005777623918322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,5120,0.09838666518529256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,3584,0.020784889658292133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,4096,0.08265688684251574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,4096,0.05720800161361694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,3584,0.05011111166742113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,3072,0.018066666192478605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,3584,0.07531378004286024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,2560,0.015915556086434256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,3072,0.043531556924184166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,2560,0.03839022252294753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,3072,0.06827821996476915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,2560,0.060066666867997914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,2048,0.013822221921549903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,1536,0.011185777684052786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,2048,0.05297688974274529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,2048,0.03309955530696445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,1536,0.026649778087933857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,1024,0.010100444157918295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,1536,0.04356088903215197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,768,0.008664000365469191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,1024,0.02162666618824005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,1024,0.03708444370163812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,512,0.007303111255168915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,768,0.03286666671435038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,768,0.018260444204012554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,256,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,512,0.02957333458794488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,256,0.014800889624489678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,512,0.01664266652531094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,128,0.004909333255555895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,256,0.02676444417900509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,64,0.00461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,3072,128,0.023011555274327595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,128,0.014565333724021912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,3072,32,0.004640000147951974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,64,0.014478221535682678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,3072,32,0.014154665999942355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,65536,0.2489546669854058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,51200,0.1926142242219713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,65536,0.8128071361117892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,65536,1.0700871149698894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,51200,0.6376088990105523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,16384,0.06464622418085735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,16384,0.20523288514879015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,51200,0.8406017621358236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,12288,0.04088355435265435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,16384,0.28268710772196454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,10240,0.038189331690470375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,12288,0.15058933364020452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,12288,0.20841599835289848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,8192,0.029648890097935993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,10240,0.17503733105129668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,10240,0.12407910823822021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,8192,0.10068355666266547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,7168,0.027120888233184814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,8192,0.14340355661180285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,6144,0.024159111910396155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,7168,0.08925422032674153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,7168,0.12833689318762884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,5120,0.020604444874657523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,6144,0.07791466845406426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,6144,0.11310756206512451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,4096,0.017376888129446242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,5120,0.06622133652369182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,5120,0.09799822171529134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,3584,0.016193777322769165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,4096,0.05310577816433377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,4096,0.08270755741331312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,3584,0.047147555483712085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,3072,0.014244443840450711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,3584,0.07503555880652533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,2560,0.012344000240166983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,3072,0.04149688945876227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,3072,0.06829510794745551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,2560,0.03652711047066583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,2048,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,2560,0.06033511294258965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,1536,0.008979555633332994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,2048,0.03163644340303209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,1536,0.025115556187099878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,1024,0.007305777735180325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,1536,0.044334222873051964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,1024,0.0207022229830424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,1024,0.03703377644220988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,768,0.006554666492674086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,768,0.03266222278277079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,768,0.01792977750301361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,512,0.005643555687533484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,512,0.01585422290696038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,256,0.004909333255555895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,256,0.0144177774588267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,2048,0.05249511202176412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,128,0.004609777695602841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,256,0.026883555783165827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,128,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,128,0.02308888898955451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,64,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2560,32,0.004243555582231945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,64,0.013834666874673633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2560,512,0.029185778564876978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2560,32,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,65536,0.1786044438680013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,51200,0.14290311601426867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,65536,0.7684453328450521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,51200,0.6034577687581381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,16384,0.047974222236209445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,65536,1.0699119567871094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,16384,0.1954408884048462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,12288,0.03502755694919162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,16384,0.28208888901604545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,12288,0.1430479950375027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,51200,0.840460459391276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,10240,0.029650668303171795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,10240,0.11785244941711426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,12288,0.2088559998406304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,10240,0.17476356029510498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,8192,0.025480889611774023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,7168,0.023899555206298828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,8192,0.1428862280315823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,7168,0.08463644319110447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,6144,0.020439111524158053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,7168,0.12799289491441515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,6144,0.07374578052096896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,5120,0.017479999197853934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,6144,0.11309244897630479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,5120,0.0616497794787089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,4096,0.015258666541841296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,5120,0.09772355688942803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,4096,0.049719111786948315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,4096,0.08229600058661567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,3584,0.014225777652528552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,3072,0.013071999781661563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,3584,0.044828444719314575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,3584,0.07457511292563544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,2560,0.01057866629627016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,3072,0.039515554904937744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,3072,0.06708888875113593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,2048,0.009721777505344814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,2560,0.03460799985461765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,2560,0.059487111038631864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,2048,0.029594666428036157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,1536,0.008028444316652086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,2048,0.052339557144376964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,1536,0.024047111471494038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,1024,0.006545777950021956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,1536,0.04309511184692383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,8192,0.09511378076341416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,1024,0.03632444474432204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,1024,0.01963200006220076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,768,0.005956444475385878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,512,0.005285333428117964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,768,0.03224622209866842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,768,0.016869333055284288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,256,0.004545777622196409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,512,0.015442666080262927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,512,0.028527110815048218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,128,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,256,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,256,0.026830222871568467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,128,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,2048,128,0.022708444131745234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,32,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,64,0.01314133322901196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,2048,32,0.013088000317414602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,65536,0.14871644973754883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,65536,0.7252844704522027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,2048,64,0.0038728887836138406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,51200,0.11089955435858832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,65536,0.9387707180447049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,16384,0.04047111007902358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,51200,0.5703493224249946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,16384,0.18617867098914254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,12288,0.030251555972629126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,51200,0.7382088767157661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,12288,0.13548089398278132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,16384,0.24983821974860299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,10240,0.024078221784697637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,12288,0.18415378199683297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,8192,0.021086222595638696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,10240,0.1546142233742608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,10240,0.11141688293880886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,8192,0.09002577596240574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,7168,0.018609777092933655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,8192,0.1262648900349935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,6144,0.016533333394262526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,7168,0.07955288887023926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,7168,0.11371821827358669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,5120,0.014594667487674289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,6144,0.06952977842754789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,6144,0.10045778089099461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,4096,0.012482666307025485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,5120,0.0867982241842482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,5120,0.05822755230797661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,3584,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,4096,0.047264890538321606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,4096,0.07370222277111478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,3072,0.010021333065297868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,3584,0.06709511412514581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,3584,0.04259999924235874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,3072,0.03709777858522203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,2560,0.00906755526860555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,3072,0.059789333078596324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,2048,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,2560,0.03189511100451151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,2560,0.05264000097910563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,2048,0.027828445037206013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,1536,0.00701155596309238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,2048,0.04585777719815572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,1024,0.005934222290913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,1536,0.023042667243215773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,1536,0.03934577769703335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,768,0.005218666460778978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,1024,0.033626665671666466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,768,0.016576889488432143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,768,0.02981333269013299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,512,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,512,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,256,0.004243555582231945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,512,0.027441778116756018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,256,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,256,0.024898666474554274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,128,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1536,128,0.021673777037196692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,1024,0.018944000204404194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,64,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1536,32,0.003604444364706675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,64,0.012828444441159567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,32,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,65536,0.11974578433566624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,65536,0.681156423356798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,51200,0.09364355272716945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1536,128,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,65536,0.9380488925509982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,16384,0.032171554035610624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,51200,0.534190230899387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,16384,0.17607200145721436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,12288,0.023792889383104112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,51200,0.7387217945522733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,16384,0.2484124501546224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,12288,0.12778755029042563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,10240,0.020417778028382193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,12288,0.18349866072336832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,8192,0.01777066621515486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,10240,0.15361332893371582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,10240,0.10480178064770168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,7168,0.015563555889659457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,8192,0.08449244499206543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,8192,0.12559910615285239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,6144,0.01420088940196567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,7168,0.07504710886213514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,7168,0.1126284466849433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,6144,0.06478577852249146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,5120,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,6144,0.09953688912921482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,5120,0.05435377690527174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,4096,0.01053066634469562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,5120,0.08706755770577325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,3584,0.009457777771684859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,4096,0.04392355680465698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,4096,0.0727671119901869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,3072,0.008447111480765873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,3584,0.039593779378467135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,3584,0.06655644708209567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,2560,0.007697777615653143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,3072,0.03472088774045309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,3072,0.06024977895948622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,2560,0.030360887447992962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,2048,0.006991110742092133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,2560,0.051763554414113365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,1536,0.0063591111037466265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,2048,0.026304889056417678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,2048,0.046169777711232506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,1024,0.00499288903342353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,1536,0.022227555513381958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,1536,0.03871022330390082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,768,0.004928888960017098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,1024,0.01828888886504703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,1024,0.033315555916892156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,768,0.015517334143320719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,512,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,512,0.014109333356221518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,768,0.03018222252527873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,512,0.026855111122131348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,256,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,128,0.003940444439649582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,256,0.024566221568319533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,256,0.013047110703256396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,128,0.012433778080675336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,64,0.003714666598372989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,1024,32,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,64,0.01239911135700014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,65536,0.08702755636639065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,1024,32,0.012463111016485425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,65536,0.6412009133232964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,51200,0.07053511010275947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,1024,128,0.022103110949198406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,65536,0.927536858452691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,16384,0.02782577938503689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,51200,0.5041155815124512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,16384,0.16703822877671984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,12288,0.022156443860795762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,51200,0.7288328806559244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,16384,0.24632000923156738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,10240,0.019281778070661757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,12288,0.12156444125705296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,12288,0.18148177199893525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,10240,0.0992444422509935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,8192,0.016189333465364244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,10240,0.15248889393276638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,8192,0.07972444428337945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,8192,0.12404355737898086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,7168,0.014351111319330005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,6144,0.0127688886390792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,7168,0.07050310903125338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,7168,0.11122043927510579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,5120,0.011153777440388998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,6144,0.06173421939214071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,6144,0.0987137754758199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,4096,0.009379555781682333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,5120,0.051375110944112144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,5120,0.08582666847440933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,3584,0.00867733359336853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,4096,0.041625777880350746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,4096,0.07207910882102118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,3072,0.007703111403518253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,3584,0.06544178062015109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,3584,0.03744711147414313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,3072,0.0332951115237342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,2560,0.007263999846246507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,3072,0.05868444177839491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,2048,0.006639110959238476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,2560,0.02847555610868666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,2560,0.05308977762858073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,2048,0.025524444050259058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,1536,0.005984000033802456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,2048,0.04519644379615784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,1024,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,1536,0.02100533412562476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,1536,0.03874933388498094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,1024,0.01756533318095737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,768,0.004592888885074192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,1024,0.03294310967127482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,768,0.015186667442321777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,768,0.029669334491093952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,512,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,512,0.013765333427323235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,512,0.026905778381559584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,256,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,256,0.02402577797571818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,256,0.01277511070171992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,128,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,128,0.012201777762836881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,768,128,0.021528888079855178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,64,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,768,32,0.00351555562681622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,64,0.012083555261294046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,65536,0.0640995568699307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,768,32,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,51200,0.05247822072770861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,65536,0.6387226846483018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,65536,0.9260604646470812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,51200,0.5015235477023655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,16384,0.022970666488011677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,16384,0.1650622288386027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,12288,0.018887110882335238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,16384,0.24592177073160806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,51200,0.7289955351087781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,10240,0.017806222041447956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,12288,0.11981066068013509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,12288,0.1811760001712375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,8192,0.013259555730554791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,10240,0.0980222225189209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,10240,0.15159555276234946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,7168,0.012390221986505719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,8192,0.07934221956464979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,8192,0.12422755029466416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,6144,0.010712888505723743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,7168,0.06995644172032674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,7168,0.11154044999016656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,6144,0.06026666694217258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,5120,0.009451555709044138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,6144,0.09794133239322239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,5120,0.0503751105732388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,5120,0.08517155382368301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,4096,0.039565334717432656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,4096,0.0716080003314548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,3584,0.007306666837798224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,3584,0.03668177790111966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,3584,0.06494844622082181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,3072,0.006711111300521427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,3072,0.03291644321547614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,3072,0.056800001197391085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,2560,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,2560,0.028404444456100464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,2048,0.005577777822812398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,4096,0.007997333175606197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,2048,0.025034666061401367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,1536,0.00488711107108328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,1536,0.020670221911536325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,1536,0.03802311089303758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,1024,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,1024,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,2560,0.05071999960475498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,768,0.004211555752489301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,1024,0.03276088833808899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,768,0.015210666590266757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,512,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,2048,0.0448000000582801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,768,0.029487109846538965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,256,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,512,0.026390223039521113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,512,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,128,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,256,0.012440889245933957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,256,0.024062222904629175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,64,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,512,128,0.022016000416543748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,128,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,512,32,0.003168888803985384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,65536,0.044806222120920815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,64,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,512,32,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,51200,0.03717155588997735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,65536,0.5998524559868706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,51200,0.4723831282721625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,16384,0.014479999740918478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,65536,0.9252711402045356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,16384,0.15664177470737034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,12288,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,16384,0.24511377016703287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,12288,0.11281066470675999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,51200,0.7286559740702311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,10240,0.01238844460911221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,8192,0.01204355557759603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,12288,0.18137421872880724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,10240,0.0922551088862949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,7168,0.011052444577217102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,8192,0.0740648905436198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,8192,0.12452533509996201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,7168,0.06523111131456164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,6144,0.010020444790522257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,7168,0.11119022634294297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,6144,0.0565155545870463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,6144,0.09781866603427464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,5120,0.00868800034125646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,5120,0.04662044511901008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,5120,0.08446311288409764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,4096,0.007359111474619971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,4096,0.036972443262736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,4096,0.07084888882107206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,10240,0.15196800231933594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,3584,0.034006221426857844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,3584,0.06331200069851346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,3072,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,3072,0.03014311194419861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,3072,0.05648710992601183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,2560,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,2560,0.026895110805829365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,2560,0.0503475566705068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,2048,0.005223111146026188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,2048,0.023704888092146978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,2048,0.04486133323775398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,1536,0.004868444469239977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,3584,0.00693866651919153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,1024,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,1536,0.03772622346878052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,1024,0.016703999704784818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,768,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,768,0.014138667119873894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,768,0.02921688887808058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,512,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,512,0.01297155519326528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,512,0.02572088936964671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,1536,0.01990133358372582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,256,0.003506666670242945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,1024,0.03254577848646376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,256,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,256,0.02365155518054962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,128,0.0030133333057165146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,64,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,256,128,0.02091022166940901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,256,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,64,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,65536,0.03557422094874912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,32,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,65536,0.5996541976928711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,51200,0.02937777837117513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,256,128,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,65536,0.9264311260647244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,16384,0.012039999994966718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,51200,0.47188531027899844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,16384,0.15618311034308538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,12288,0.010320000350475311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,16384,0.2457182142469618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,51200,0.7280684577094184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,12288,0.11247821648915608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,10240,0.009674666656388177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,12288,0.18121066358354357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,8192,0.008663111262851292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,10240,0.09171644184324478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,10240,0.15135467052459717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,7168,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,8192,0.07381066348817614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,8192,0.12374222278594971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,6144,0.008264888491895463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,7168,0.06520355410046048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,7168,0.11016088724136353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,5120,0.007630222373538547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,6144,0.05561155743069119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,6144,0.09724089172151353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,4096,0.007321777443091075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,5120,0.04594044552909004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,5120,0.08414133389790852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,3584,0.006959999601046245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,4096,0.037124445041020714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,4096,0.06902133093939887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,3072,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,3584,0.03357866737577651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,3584,0.0631928907500373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,2560,0.005567111074924469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,3072,0.030238221089045208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,3072,0.05746399694018894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,2048,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,2560,0.026751110951105755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,2560,0.05069066749678718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,1536,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,2048,0.02360533343421088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,2048,0.04454311066203647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,1024,0.0041964443193541635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,1536,0.038027554750442505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,768,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,1024,0.03260711166593764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,1536,0.020350222786267597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,768,0.014463111758232117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,768,0.029222223493787978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,512,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,256,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,512,0.025765332910749648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,512,0.01308444473478529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,256,0.023769777682092454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,128,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,256,0.012126222252845764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,128,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,64,0.00282222218811512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1280,128,128,0.020942222740915086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,128,32,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,1024,0.016891555653678048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,64,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,65536,0.033896889951494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,51200,0.02779200010829502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,128,32,0.01181777815024058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,16384,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,65536,0.5999893082512749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,51200,0.47136979632907444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,16384,0.1524346669514974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,12288,0.009703999592198266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,10240,0.008996444443861643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,8192,0.008386666576067606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,12288,0.11243377791510688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,10240,0.09149599737591213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,7168,0.007986666427718269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,6144,0.007278222176763747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,8192,0.07372711102167766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,5120,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,7168,0.06491288873884413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,4096,0.007310222420427535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,6144,0.05578311284383138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,3584,0.006899555524190267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,5120,0.04632444514168633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,4096,0.03724622064166599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,3072,0.006329777754015393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,3584,0.03328088919321696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,2560,0.005934222290913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,3072,0.03037422233157688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,2048,0.005333333379692501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,2560,0.02677333354949951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,1536,0.004605333424276776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,1024,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,2048,0.023412444525294836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,768,0.0036026665733920205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,1536,0.01996799972322252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,512,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,1024,0.01687733332316081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,768,0.014202665951516895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,512,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,128,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,256,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,64,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,128,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,32,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,64,0.011153777440388998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,65536,0.03122577733463711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,64,32,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,51200,0.026196445027987163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,64,256,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,16384,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,65536,0.600164466434055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,51200,0.4711306889851888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,12288,0.1120720042122735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,16384,0.15614222155676946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,10240,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,8192,0.00942133367061615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,8192,0.07348444726732042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,10240,0.09137155612309773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,7168,0.00903200027015474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,6144,0.008404444489214156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,7168,0.06516977813508776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,6144,0.05563644568125407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,12288,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,5120,0.007707555260923173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,4096,0.007649777664078607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,3584,0.006672888994216919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,5120,0.04562044474813673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,3072,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,4096,0.0370488895310296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,3584,0.0329733325375451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,2560,0.005664888769388199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,2048,0.005096000101831225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,3072,0.030439999368455675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,1536,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,2560,0.02640977833006117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,1024,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,1536,0.019827556278970506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,2048,0.0237137774626414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,768,0.003531555748648114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,1024,0.016538666354285348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,512,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,256,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,512,0.013078221844302284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,128,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,256,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,768,0.014520888527234396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,64,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1280,32,32,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,128,0.011218667030334473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,64,0.011529777612951068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1280,32,32,0.01165422217713462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,16384,1.1551262537638347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,16384,1.4579057693481445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,51200,3.4886355929904513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,12288,0.8729671372307671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,10240,0.7284151183234321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,12288,1.0568781958685982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,51200,5.039624955919054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,8192,0.5928959846496582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,10240,0.8632026778327094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,8192,0.6942390865749783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,7168,0.5214311281840006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,7168,0.5952995618184408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,6144,0.45291731092664933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,6144,0.5141066445244683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,5120,0.38377422756618923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,4096,0.30916089481777614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,5120,0.4278675450219049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,4096,0.3466240035163031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,3584,0.27274489402770996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,3072,0.23653690020243326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,3072,0.26029245058695477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,2560,0.1991786691877577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,3584,0.3029635482364231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,2048,0.2132862276501126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,2560,0.22329865561591256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,2048,0.1834017833073934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,1536,0.14572532971700033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,1024,0.11582844787173802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,768,0.09324533409542507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,1024,0.11039644479751587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,512,0.06323911084069146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,768,0.09021155701743232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,256,0.039818667703204684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,512,0.07835733228259616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,128,0.0316880014207628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,256,0.07191199726528592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,128,0.06757689184612699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,1536,0.16342400179968938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,32,0.030590222941504583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,64,0.06736355357699923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,65536,32,0.0674275557200114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,65536,64,0.029304001066419814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,65536,3.4859511057535806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,51200,2.7414266798231335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,16384,0.8911280102199979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,65536,5.021430969238281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,16384,1.1776267157660592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,12288,0.6765848795572916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,10240,0.5709724426269531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,12288,0.8355457517835828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,10240,0.6916275554233127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,8192,0.46163108613755965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,7168,0.40730932023790145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,8192,0.5413582060072157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,7168,0.47997686598036027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,6144,0.3529173268212213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,51200,3.8556399875217013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,6144,0.4127262168460422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,4096,0.24130309952629936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,5120,0.3410702281528049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,3584,0.21249422762129042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,3072,0.18398222658369276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,4096,0.27738221486409503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,2560,0.15730488300323486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,3072,0.2097111145655314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,2048,0.13004622194502088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,5120,0.29900532298617893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,1536,0.10247466961542766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,2048,0.14786844783359104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,1024,0.07197511196136475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,1536,0.11908622582753499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,768,0.05863022141986423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,3584,0.24031821886698404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,1024,0.08972799777984619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,512,0.04407377706633674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,256,0.03179644544919332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,768,0.07391999827490912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,512,0.06442044178644817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,256,0.05892799960242378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,64,0.02130222154988183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,2560,0.18060889508989122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,32,0.021595555875036452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,128,0.05545511179500156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,64,0.05551822317971123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,51200,32,0.055677334467569985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,51200,128,0.024294222394625347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,51200,0.8075315687391492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,65536,1.1722515953911674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,16384,0.2640053431193034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,12288,0.19530488385094535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,16384,0.4309431182013617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,65536,1.8018364376491969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,10240,0.2575937906901042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,12288,0.3121031125386556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,8192,0.20850133895874023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,10240,0.25946044921875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,51200,1.4306035571628148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,8192,0.20840977297888863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,6144,0.14854221873813206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,7168,0.1855591138203939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,5120,0.1252711084153917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,6144,0.16108889049953884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,4096,0.10014844603008694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,5120,0.1365573273764716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,3584,0.08898311191134983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,4096,0.11201777723100449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,3072,0.07509511046939425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,3584,0.0992657807138231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,2560,0.0655004448360867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,7168,0.17976978090074328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,3072,0.08808622095319961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,2048,0.053708444039026894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,2560,0.0766400032573276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,1536,0.04152089026239183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,1024,0.028841776980294123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,2048,0.06366933054394193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,1536,0.05120799938837687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,768,0.02273955610063341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,1024,0.03901244534386529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,512,0.01551555593808492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,256,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,768,0.032630221711264715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,128,0.010123555858929953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,512,0.02790844440460205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,64,0.009382222261693742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,256,0.025103999508751765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,128,0.024104000793562994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,16384,32,0.009379555781682333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,64,0.0240666667620341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,16384,32,0.023652444283167522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,65536,0.7635431289672852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,51200,0.595512866973877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,16384,0.1992808845308092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,16384,0.3482195536295573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,65536,1.4636444515652125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,12288,0.1516337792078654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,10240,0.12831377983093262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,12288,0.2542160087161594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,51200,1.1498212814331055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,8192,0.10480266809463501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,7168,0.0926026635699802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,10240,0.2119751109017266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,8192,0.17197510931226942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,6144,0.08102133538987902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,5120,0.06902044349246554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,7168,0.1514213350084093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,4096,0.05803022119734022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,6144,0.13206311066945395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,5120,0.11291821797688802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,3584,0.05024088753594292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,3072,0.044417778650919594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,4096,0.0927013357480367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,3584,0.08271644512812297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,2560,0.040045334233178034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,2048,0.032144000132878624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,3072,0.07374133004082574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,1536,0.02492800023820665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,2560,0.06389333142174615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,1024,0.01834488908449809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,2048,0.051938666237725147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,768,0.015070221490330167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,1024,0.032560000816980995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,768,0.028010666370391846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,512,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,1536,0.043266667260064014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,256,0.009649777577983009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,128,0.008016000191370646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,512,0.023444443941116333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,64,0.006923555499977536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,128,0.021031111478805542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,12288,32,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,64,0.020622221959961783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,32,0.020662221643659804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,65536,0.7620862325032552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,12288,256,0.022091555926534865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,65536,1.2748746871948242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,16384,0.19404710663689506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,51200,1.010984844631619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,16384,0.3063804573482937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,12288,0.1476915544933743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,10240,0.1482328838772244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,12288,0.2250293360816108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,51200,0.599230236477322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,8192,0.11870843834347194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,7168,0.10304088724984063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,10240,0.18806844287448457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,8192,0.15170310603247747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,6144,0.08911466598510742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,7168,0.13458666536543104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,5120,0.07472533649868436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,6144,0.11746488677130805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,4096,0.06069778071509468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,5120,0.09967288706037734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,3584,0.05437155564626058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,3072,0.0470124449994829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,4096,0.08293066422144572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,2560,0.04000533289379544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,3584,0.07416533099280463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,2048,0.032968888680140175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,2560,0.05607022179497612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,1536,0.02446311049991184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,2048,0.0462719996770223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,3072,0.06567911307017009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,1024,0.017986666825082567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,768,0.014535110857751636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,1536,0.038047999143600464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,512,0.01126399967405531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,1024,0.02966044346491496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,768,0.025615110993385315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,512,0.021632888250880774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,256,0.008416000339719985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,128,0.007958222594526079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,64,0.006631999793979857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,256,0.019996444384257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,128,0.01961688862906562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,10240,32,0.006776888751321369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,64,0.019285332825448778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,10240,32,0.019296000401178997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,65536,0.8147813479105631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,51200,0.6382817692226833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,65536,1.114154709710015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,16384,0.267215993669298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,51200,0.8722097608778211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,12288,0.15879733032650417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,10240,0.097744001282586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,12288,0.19599644343058267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,8192,0.0776711106300354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,16384,0.20937777890099418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,10240,0.16429244147406682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,7168,0.0702986650996738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,6144,0.061391108565860324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,8192,0.13335466384887695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,5120,0.051240887906816274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,6144,0.10435022248162164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,5120,0.08875288565953572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,4096,0.04307555490069919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,3584,0.038712001509136625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,4096,0.07242666350470649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,3072,0.03397511111365424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,3584,0.06591644552018908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,2560,0.02971822354528639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,3072,0.05755911270777384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,2048,0.02481866710715824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,7168,0.11856177118089463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,1536,0.02071555621094174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,2560,0.05048000150256687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,2048,0.04130488965246413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,1536,0.0346684455871582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,768,0.01276622215906779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,1024,0.027178666657871667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,512,0.010426667001512317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,768,0.023102222217453852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,256,0.008044444852405125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,512,0.01993600030740102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,128,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,256,0.01826844447188907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,64,0.0063031112982167145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,128,0.01798311041461097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,32,0.006353777729802662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,64,0.017550221747822232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,8192,32,0.017623111605644226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,65536,0.468231995900472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,51200,0.3634782102372911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,65536,1.0219430923461914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,8192,1024,0.016167110866970487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,16384,0.11948800086975098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,12288,0.0904995534155104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,51200,0.8142702314588758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,16384,0.24917954868740508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,12288,0.1825395557615492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,10240,0.07795111338297527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,8192,0.0653324458334181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,7168,0.05537688732147217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,10240,0.1527200010087755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,8192,0.12399910555945502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,6144,0.04741955465740628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,7168,0.11027022202809651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,5120,0.041069332096311785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,4096,0.034259554412629865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,6144,0.09668533007303874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,3584,0.03070311082734002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,5120,0.0834088921546936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,3072,0.027293332748942908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,4096,0.06862844361199273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,2560,0.023668444818920557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,3072,0.05347911185688443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,2048,0.02032177812523312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,2560,0.04692711101637947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,1536,0.016536888149049547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,2048,0.03882755504714118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,1024,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,1536,0.0329715543323093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,768,0.010816888676749336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,1024,0.025387555360794067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,512,0.009739555418491364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,768,0.02200088898340861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,512,0.01889600025282966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,256,0.008026666939258575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,128,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,3584,0.05865777863396538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,64,0.005294222384691238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,256,0.01756800048881107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,7168,32,0.005744888964626525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,64,0.016881777180565726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,32,0.016493333710564505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,65536,0.40159111552768284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,51200,0.3022844526502821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,7168,128,0.017408889200952318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,16384,0.10112622049119736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,51200,0.7494791348775228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,12288,0.07856177621417575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,12288,0.16767199834187826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,10240,0.06763733095592923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,65536,0.9396826426188151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,8192,0.056391113334231906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,10240,0.14056621657477483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,7168,0.04816888769467672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,8192,0.11437777678171794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,6144,0.04333066609170702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,16384,0.22971110873752168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,7168,0.10258755418989395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,5120,0.036128888527552284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,4096,0.02977333466211955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,6144,0.09018044339285956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,3584,0.027775999572541978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,5120,0.07645422220230103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,3072,0.0236151119073232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,4096,0.06296710835562812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,3584,0.054811557133992515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,3072,0.048208889034059316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,2560,0.021783111823929682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,2048,0.018218666315078735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,1536,0.014831110835075378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,2560,0.04292000002331204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,1024,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,1536,0.030252443419562444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,2048,0.036517331997553505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,768,0.009128889275921715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,1024,0.023256889647907678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,512,0.0073928890956772705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,256,0.006266666783226862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,768,0.020971554848882887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,512,0.018193778064515855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,256,0.016771554946899414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,128,0.005608888963858287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,64,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,6144,32,0.0053004444473319584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,128,0.016196444630622864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,64,0.01625066664483812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,6144,32,0.01552977826860216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,65536,0.3510631190405951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,51200,0.28192800945705837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,16384,0.09440888961156209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,16384,0.20999822351667616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,65536,0.8583306736416287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,12288,0.07275822427537706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,51200,0.6883804533216688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,10240,0.07601866457197401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,8192,0.06150488720999824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,10240,0.1290960046980116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,12288,0.15333955817752415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,7168,0.0544088880221049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,8192,0.10576799843046401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,6144,0.047657777865727745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,5120,0.040024889840020075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,6144,0.08258044719696045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,4096,0.03193688723776076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,5120,0.07020444340176053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,3584,0.028041776683595445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,4096,0.05771911144256592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,3072,0.024860444996092055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,3584,0.05024977856212192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,2560,0.021243555678261652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,3072,0.044863111442989774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,2048,0.01795644395881229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,2560,0.03976888789070977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,1536,0.014465777410401238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,2048,0.0332568883895874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,1024,0.010863111250930362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,1536,0.028288000159793433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,768,0.009085333181752099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,7168,0.09471199909845988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,1024,0.022043555974960327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,768,0.019648889700571697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,256,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,512,0.016881777180565726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,128,0.005232000102599462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,256,0.01588888963063558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,64,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,128,0.015512888630231222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,32,0.005290666802061929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,64,0.015217777755525378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,5120,32,0.014872888724009195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,65536,0.32786311043633354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,5120,512,0.007248000138335758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,51200,0.25378134515550405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,16384,0.08645600080490112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,65536,0.7854888704088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,12288,0.06676089101367526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,16384,0.1935795545578003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,10240,0.056800888644324414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,51200,0.6171902020772299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,8192,0.045999109745025635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,12288,0.14226755830976698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,7168,0.0403404434521993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,10240,0.1199902163611518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,8192,0.09885155492358738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,6144,0.03462844424777561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,7168,0.08799555566575791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,5120,0.02906399965286255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,4096,0.02382844520939721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,6144,0.07678844531377156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,3584,0.021556443638271753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,5120,0.06520533561706543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,3072,0.0188737776544359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,4096,0.0536480016178555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,2560,0.016591999265882704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,3584,0.04691733254326714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,2048,0.014402666025691561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,3072,0.041899555259280734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,1536,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,2560,0.03774311145146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,2048,0.03086400032043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,1024,0.009151111046473185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,1536,0.025777777036031086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,768,0.007679111427730984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,512,0.006667555620272954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,1024,0.020377778344684176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,256,0.005550222264395819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,768,0.018363555272420246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,128,0.004866666677925321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,512,0.016347555650605094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,64,0.004950222041871813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,256,0.014889778362380134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,128,0.014434667097197639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,4096,32,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,64,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,4096,32,0.01407199932469262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,65536,0.25034666061401367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,51200,0.19735822412702772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,16384,0.06821688678529528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,16384,0.1833111047744751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,65536,0.7449430889553494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,51200,0.583204428354899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,12288,0.05145600106981066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,10240,0.04493155413203769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,8192,0.036891556448406644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,12288,0.135535995165507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,10240,0.11425244808197021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,7168,0.03129511078198751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,8192,0.09403200281990899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,6144,0.02702755563788944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,5120,0.022851554883850947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,7168,0.08348177538977729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,4096,0.019335110982259113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,6144,0.07292266686757405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,3584,0.01779022150569492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,4096,0.05032622151904636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,3072,0.015561777684423657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,3584,0.04493155413203769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,2560,0.014262222581439547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,3072,0.04007822275161743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,2048,0.011974222130245633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,2560,0.03541510966089036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,1536,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,2048,0.029517332712809246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,1024,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,1536,0.024903999434577093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,768,0.0069955554273393415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,1024,0.019609777463806998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,512,0.005894222193294101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,768,0.017830222845077515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,256,0.005190222213665645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,5120,0.061977777216169566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,512,0.015819556183285184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,128,0.004519111166397731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,64,0.004035555654101902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,256,0.014535110857751636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3584,32,0.004265777766704559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,128,0.014110222458839417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,64,0.014087999860445658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3584,32,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,65536,0.20823378033108184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,51200,0.1605679988861084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,16384,0.05599821938408745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,16384,0.17467555734846327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,65536,0.7035644319322375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,51200,0.5561884244283041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,12288,0.04212177793184916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,10240,0.03794222076733907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,8192,0.030530667967266504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,12288,0.1281928883658515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,10240,0.10844177669949001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,7168,0.02553333342075348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,8192,0.08892444769541423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,6144,0.022802667485343084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,5120,0.019880000087949965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,7168,0.07902310954199897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,4096,0.01737777723206414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,6144,0.06875110997094049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,3584,0.015581332974963717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,5120,0.058176000912984215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,4096,0.04701688885688782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,3072,0.014162666267818876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,3584,0.04192355606291029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,2560,0.01239911135700014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,3072,0.03804088963402642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,2048,0.011037333144081963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,2560,0.03352533446417914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,1536,0.008745777938101027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,2048,0.028299556838141546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,1024,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,1536,0.023727110690540735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,768,0.006224888894293044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,1024,0.018940443793932598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,512,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,768,0.01716977854569753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,256,0.0046817776229646476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,512,0.015484443969196744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,128,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,128,0.013463111387358772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,64,0.0038151111867692736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,64,0.013422222601042854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,3072,32,0.003863999827040566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,32,0.013467555244763693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,65536,0.20294488800896537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,3072,256,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,51200,0.15850399600134954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,16384,0.05506489011976454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,65536,0.6629244486490885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,16384,0.16469777954949275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,12288,0.04278577698601616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,51200,0.5200444327460395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,10240,0.03571110963821411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,12288,0.12122221787770589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,8192,0.028130667077170476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,10240,0.10208444462882148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,8192,0.08376266558965047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,6144,0.02141777839925554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,7168,0.07465866539213392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,5120,0.018565333551830716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,6144,0.0648533304532369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,4096,0.01588355501492818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,5120,0.05452355411317614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,3584,0.01423377792040507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,4096,0.04422666629155477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,3072,0.013173333472675748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,3072,0.03601422243648105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,7168,0.02477600011560652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,2560,0.01166311071978675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,2048,0.010058666268984476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,2560,0.032390223609076604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,2048,0.026458667384253606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,1536,0.00869066682126787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,1024,0.006745777610275481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,1536,0.02271911170747545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,768,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,768,0.016561778055297006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,1024,0.018351111147138808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,3584,0.04011733333269755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,512,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,256,0.0046160001721647055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,128,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,256,0.01368533323208491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,64,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,128,0.013501333693663279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,512,0.015180443723996481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2560,32,0.004264889078007804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,64,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,65536,0.1821697817908393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2560,32,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,51200,0.14419821898142496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,16384,0.04997422297795614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,65536,0.6176586680942112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,16384,0.15471377637651232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,51200,0.487368901570638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,12288,0.03809955716133118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,10240,0.03174933460023668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,8192,0.02604177759753333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,12288,0.11351199944814046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,7168,0.02293777796957228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,10240,0.09606755442089504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,6144,0.0199297782447603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,8192,0.07879733377032809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,7168,0.06988000207477145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,5120,0.01757777730623881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,6144,0.06043644746144613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,4096,0.014892444014549255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,3584,0.01352266636159685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,5120,0.05130755570199755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,4096,0.04152089026239183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,3584,0.037957333856158786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,2560,0.010602666272057427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,3072,0.03359199894799127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,2048,0.00906755526860555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,2560,0.03026755650838216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,2048,0.025087111526065405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,1536,0.007690666450394525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,1024,0.006143110907740063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,1536,0.021686222818162706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,768,0.005564444594913059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,1024,0.017535999417304993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,768,0.015862221519152325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,512,0.004930666751331753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,3072,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,256,0.004206222378545337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,512,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,128,0.004182222402758068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,256,0.01347644461525811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,64,0.0038782221575578055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,128,0.013134222063753339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,2048,32,0.004201777693298128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,64,0.012871999707486896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,2048,32,0.012507555385430654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,65536,0.15564000606536865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,51200,0.12309777736663818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,16384,0.04281244344181485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,51200,0.4537342389424642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,65536,0.5774701966179742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,16384,0.1453582180870904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,10240,0.027560888065232172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,12288,0.10642311308119033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,8192,0.023056000471115112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,10240,0.08947377734714085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,7168,0.02026844521363576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,8192,0.073381331231859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,6144,0.018230222993426852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,7168,0.06510488854514228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,5120,0.015609777635998197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,6144,0.057096891933017306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,4096,0.013562666873137156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,5120,0.04764444298214383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,12288,0.03248000144958496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,3584,0.01186044431395001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,4096,0.0387644436624315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,3072,0.010468444062603844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,2560,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,3584,0.03492622243033515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,2048,0.007657777931955125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,3072,0.031108445591396753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,1536,0.007044444481531779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,2560,0.027943111128277246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,1024,0.006021333237489064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,2048,0.02368444369898902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,768,0.0053288886944452924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,1536,0.020375111036830477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,512,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,1024,0.01686311099264357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,256,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,768,0.015216888652907478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,128,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,512,0.013884444203641681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,64,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,256,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1536,32,0.0035013332962989807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,64,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,32,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,65536,0.13233333163791233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,51200,0.09538488917880589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,65536,0.5351066589355469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,51200,0.42017510202195907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,16384,0.03124888737996419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,12288,0.023697778582572937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1536,128,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,16384,0.13483822345733643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,10240,0.02217866645918952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,12288,0.09910133149888779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,8192,0.018608889646000333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,7168,0.01638488968213399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,10240,0.08328000042173597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,8192,0.06782666842142741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,6144,0.014533332652515836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,7168,0.059934225347306996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,5120,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,4096,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,6144,0.051107555627822876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,3584,0.009201777478059133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,4096,0.03543022274971008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,3072,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,3584,0.03247200118170844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,2560,0.007629333270920648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,3072,0.02869866622818841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,2048,0.0069413334131240845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,2560,0.026473777161704168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,1536,0.005980444451173146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,2048,0.022317333353890315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,1024,0.005340444544951121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,1536,0.01958933307064904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,768,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,1024,0.015519110692871941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,512,0.004292444636424382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,768,0.01479200025399526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,256,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,512,0.013167111410035027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,128,0.003536888708670934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,256,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,64,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,128,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,1024,32,0.003415999934077263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,64,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,32,0.01185066666867998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,65536,0.0776488913430108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,65536,0.5308080249362522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,1024,5120,0.04342400034268697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,51200,0.06316533353593615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,16384,0.02596355477968852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,16384,0.13416266441345215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,51200,0.41755376921759707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,12288,0.02541333271397485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,10240,0.021310221817758348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,12288,0.09779733419418335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,8192,0.017882666654056974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,10240,0.0825608902507358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,7168,0.016413332687483895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,8192,0.06700177987416585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,6144,0.013809777796268463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,7168,0.05889155467351278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,5120,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,6144,0.05079022381040785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,4096,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,5120,0.042565332518683545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,3584,0.010018666585286459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,4096,0.03545422355333964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,3072,0.00867555538813273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,3584,0.032250665956073336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,2560,0.007875555919276344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,3072,0.02901600135697259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,2048,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,2560,0.025711999999152288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,1536,0.005931555396980709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,2048,0.022296888960732356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,1024,0.004922666483455234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,1536,0.019267555740144517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,768,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,1024,0.015504888362354703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,512,0.003858666867017746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,768,0.014464888307783337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,256,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,512,0.013076444466908773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,128,0.0032995556377702286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,256,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,64,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,128,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,768,32,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,64,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,768,32,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,65536,0.06750133302476671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,51200,0.05482666691144308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,16384,0.020642666353119742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,65536,0.4923066563076443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,12288,0.018245332770877414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,16384,0.12417422400580512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,10240,0.016480889585283067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,51200,0.3878648810916477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,8192,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,12288,0.09083733293745254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,10240,0.07622933387756348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,7168,0.012421333127551608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,6144,0.011370666325092316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,8192,0.06209511227077908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,6144,0.04589244392183092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,5120,0.009728888670603434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,4096,0.008329778081840938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,5120,0.03879822293917338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,3584,0.008280889027648503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,4096,0.03293066554599338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,3072,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,3584,0.030076444149017334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,2560,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,2560,0.0244195560614268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,3072,0.027069333526823256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,2048,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,1536,0.005632888939645555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,7168,0.05481688843833076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,1024,0.004260444392760594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,1536,0.01847999956872728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,2048,0.02106933295726776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,768,0.0042462220622433555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,512,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,1024,0.014866666661368476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,256,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,768,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,512,0.012798222402731577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,128,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,64,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,256,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,512,32,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,128,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,65536,0.040538665321138166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,64,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,512,32,0.011175110936164856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,51200,0.033415112230512835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,16384,0.013567999833159976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,65536,0.49187999301486546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,51200,0.3867777718438043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,16384,0.12352622879876031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,12288,0.012177777787049612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,10240,0.011468444433477191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,8192,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,12288,0.08930399682786729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,7168,0.012774222426944308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,10240,0.07556711302863227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,8192,0.06100889046986898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,7168,0.05405511127577888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,6144,0.010399999717871347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,5120,0.0090844440791342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,6144,0.0453324450386895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,4096,0.00942844483587477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,5120,0.038415110773510404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,3584,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,4096,0.0321066677570343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,3072,0.0076248885856734375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,3584,0.029634667767418757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,3072,0.026103110777007207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,2560,0.005951111101441913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,2048,0.0053386667536364655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,2560,0.023974221613672044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,1536,0.004963555683692296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,2048,0.020994666549894545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,1024,0.004223111189073986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,1024,0.0151591118839052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,768,0.003976888954639435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,512,0.003738666574160258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,768,0.01421333352724711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,256,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,512,0.01292799992693795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,128,0.003000000077817175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,256,0.012051555845472546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,1536,0.018221333622932434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,64,0.0029120000286234748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,128,0.011648889217111798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,256,32,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,65536,0.03119822343190511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,64,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,256,32,0.011503111157152386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,51200,0.02741866641574436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,16384,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,16384,0.12300711207919651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,12288,0.010017777482668558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,51200,0.3867724471622043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,10240,0.011137777732478248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,12288,0.08920800023608738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,8192,0.009742221898502773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,10240,0.0749982198079427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,7168,0.009755555954244401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,8192,0.06063555346594917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,6144,0.00866844422287411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,7168,0.052689777480231396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,5120,0.008394666843944127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,6144,0.045216888189315796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,4096,0.007673777639865875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,5120,0.03843911157713996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,65536,0.4916239844428168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,3584,0.0069439998931354946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,4096,0.03180355495876736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,3584,0.02959644463327196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,3072,0.02646044393380483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,2560,0.005941333456171884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,2048,0.005296000176005893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,2560,0.024127110838890076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,1536,0.0052782222628593445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,2048,0.02103022237618764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,1536,0.01804266704453362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,1024,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,768,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,1024,0.014673777752452426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,512,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,3072,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,256,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,768,0.014198222094111972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,512,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,128,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,256,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,128,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,32,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,64,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,65536,0.029783109823862713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,128,32,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,51200,0.022297778063350256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,65536,0.4913662274678548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,16384,0.009369778136412302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,128,64,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,16384,0.12334844801161025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,12288,0.08929066525565253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,10240,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,10240,0.07473599910736084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,8192,0.009728000395827824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,51200,0.3860479990641276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,7168,0.009045333498054082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,8192,0.06094577577379015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,6144,0.009374221993817223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,7168,0.05340622199906243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,12288,0.009714666340086196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,5120,0.008380444513426887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,4096,0.007296000089910295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,6144,0.04496177699830797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,3584,0.006960888703664144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,5120,0.03860977623197768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,3072,0.006370666540331311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,3584,0.029367112451129492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,2560,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,3072,0.026533333791626826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,2560,0.02427644365363651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,2048,0.004910222358173794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,2048,0.020623111062579684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,1536,0.004272888931963178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,1536,0.018251554833518136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,1024,0.004010666575696734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,4096,0.03161599900987413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,768,0.0037439999481042228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,512,0.0033484444850020935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,1024,0.014839111102951897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,256,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,768,0.013678222066826291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,512,0.012476444244384766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,64,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,256,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,64,32,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,64,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,128,0.011547555526097616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,65536,0.02613688839806451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,64,32,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,51200,0.01758044461409251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,16384,0.012823999755912356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,16384,0.12246044476826985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,65536,0.4906284544203017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,12288,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,51200,0.3863804340362549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,10240,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,8192,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,12288,0.08887910842895508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,10240,0.0751368867026435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,7168,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,8192,0.0603662199444241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,6144,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,5120,0.010024888647927178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,7168,0.05264266663127475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,4096,0.00832799987660514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,6144,0.04517777760823568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,3584,0.007698666718271043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,5120,0.037556442949506975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,3072,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,4096,0.031904889477623835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,2560,0.006321777900060018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,3584,0.028924445311228435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,2048,0.005559999909665849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,3072,0.026133333643277485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,1536,0.004930666751331753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,2560,0.0240346673462126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,1024,0.0042168887125121225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,2048,0.020679111282030743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,768,0.0037804444630940757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,1536,0.01791733337773217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,512,0.003506666670242945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,1024,0.014825777875052558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,768,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,256,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,128,0.002943111169669363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,512,0.012484444512261285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,64,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,256,0.011704888608720569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,128,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1025,32,32,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,64,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1025,32,32,0.011154666543006897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,51200,3.061469395955404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,16384,0.9997724956936307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,51200,4.5831074184841585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,16384,1.357378641764323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,12288,0.7482186423407661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,16384,1.0535252888997395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,12288,0.9840933481852213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,10240,0.6327599949306911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,10240,0.8109688758850098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,51200,3.371035469902886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,8192,0.48784621556599933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,12288,0.7973039944966634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,8192,0.6513742340935601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,7168,0.47256798214382595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,10240,0.6726746559143066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,7168,0.5750560230678982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,6144,0.3977973461151123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,8192,0.5511404673258463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,6144,0.4924977620442708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,5120,0.31012354956732857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,7168,0.48479641808403867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,5120,0.4155866569942898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,4096,0.27901776631673175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,6144,0.42058576477898496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,4096,0.3227457735273573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,5120,0.3586311075422499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,4096,0.2954817877875434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,3584,0.28332800335354275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,3072,0.21056889163123238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,3584,0.26387911372714573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,3072,0.2475688987308078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,2560,0.17095288965437147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,3072,0.23412267367045084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,2560,0.21376444233788383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,2048,0.14352799786461726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,2560,0.20202133390638563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,2048,0.17553777164883086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,1536,0.10514577892091538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,2048,0.17032978269788954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,1536,0.1398515568839179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,1024,0.07248977820078532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,1536,0.14027021990882024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,3584,0.2451475461324056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,1024,0.10633866654502021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,768,0.05959910816616482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,1024,0.11009155379401313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,512,0.046089778343836464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,768,0.0867955552207099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,768,0.09309689203898112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,256,0.033675554725858904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,512,0.07935200134913127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,512,0.07570755481719971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,128,0.028775999943415325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,256,0.06876533561282687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,256,0.06632266441980998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,64,0.02499377727508545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,65536,128,0.05494755506515503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,128,0.06452711423238118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,65536,32,0.026123555170165166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,64,0.06446399953630236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,65536,32,0.06475910875532362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,65536,3.2786390516493054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,65536,4.553155687120226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,51200,2.575983895195855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,51200,3.6902826097276478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,65536,3.5291769239637585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,16384,0.8291849030388726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,16384,1.0784311294555664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,12288,0.6400933265686035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,12288,0.7825244267781576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,10240,0.4742710855272081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,16384,0.885651535458035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,12288,0.668050660027398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,10240,0.6430853207906088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,8192,0.3918924331665039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,8192,0.5107830895317925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,10240,0.563760863410102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,7168,0.33394845326741535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,51200,2.7646844651963978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,7168,0.4495671060350206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,6144,0.3111857838100857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,8192,0.45721064673529727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,5120,0.2539386749267578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,6144,0.3881377908918593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,7168,0.40537510977851016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,5120,0.3231075604756673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,4096,0.21083377467261422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,6144,0.3529208766089545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,3584,0.1942453384399414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,4096,0.26028532452053493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,5120,0.3012231190999349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,4096,0.25042488839891225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,3584,0.2290604379442003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,3072,0.15559021631876627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,3584,0.22234045134650338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,3072,0.20023466481102836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,3072,0.1978586647245619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,2560,0.17325333754221597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,2048,0.11508177386389838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,2560,0.17061866654290092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,1536,0.09059555663002862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,2048,0.14240088727739122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,2048,0.1440755526224772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,1536,0.11379199557834202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,1024,0.061193777455223926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,1536,0.1192622184753418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,1024,0.08656444152196248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,768,0.04887911015086704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,1024,0.0939448873202006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,2560,0.13102577792273626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,512,0.036590221855375506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,768,0.07109777794943915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,768,0.08050222529305352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,256,0.02685244381427765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,512,0.06869155830807157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,128,0.022343110707071092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,512,0.06255288918813069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,256,0.056383113066355385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,64,0.027747554911507502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,128,0.05332711007859972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,51200,32,0.029297779003779095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,64,0.053026666243871055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,32,0.05307999915546841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,51200,256,0.05695022477043999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,65536,1.0993341869778104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,51200,128,0.04785866538683573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,65536,1.6970240275065105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,51200,0.8558355437384711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,16384,0.28670401043362087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,51200,1.334930631849501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,16384,0.406805329852634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,12288,0.21704088317023384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,16384,0.39843111568027073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,12288,0.296049780315823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,10240,0.18605866697099474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,65536,1.5557884640163844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,12288,0.3013884491390652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,10240,0.2461786535051134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,51200,1.222637282477485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,8192,0.1485253307554457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,10240,0.25507111019558376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,8192,0.1990142265955607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,7168,0.13400977187686497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,7168,0.17642933792538115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,6144,0.10827555259068807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,7168,0.18714577621883818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,6144,0.15390933884514704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,6144,0.1644702222612169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,5120,0.09046666489707099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,8192,0.20933777756161162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,5120,0.13118578328026667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,4096,0.08077155881457858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,5120,0.14088800218370226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,3584,0.07163021961847942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,4096,0.11851467026604547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,4096,0.10789689090516831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,3584,0.09564089112811619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,3072,0.06067022350099352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,3584,0.10716799894968669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,2560,0.05339822173118591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,3072,0.08476622237099542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,3072,0.0972906682226393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,2048,0.04301866557863024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,2560,0.07395022445254855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,2560,0.08515555328792995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,1536,0.03055822186999851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,2048,0.07331822315851848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,2048,0.06167022387186686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,1536,0.0491848886013031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,1024,0.01975555552376641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,1536,0.06286222404903836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,768,0.01811644434928894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,1024,0.03734755516052246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,1024,0.051764445172415845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,512,0.015296889675988091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,768,0.04500444398985969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,512,0.040037333965301514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,512,0.028243554963005915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,256,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,256,0.03520444366667006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,128,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,256,0.025430222352345783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,128,0.023431110713216994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,64,0.010379555324713389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,16384,128,0.030968000491460163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,16384,32,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,768,0.031268444326188825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,64,0.023312888211674158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,16384,32,0.023391111029518977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,65536,0.8508328861660428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,65536,1.3722089131673176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,51200,0.6722462442186145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,51200,1.0908382203843858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,16384,0.21297866768307158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,65536,1.4246240191989477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,16384,0.33074845208062065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,12288,0.16720355881585014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,16384,0.36550577481587726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,51200,1.1192080179850261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,10240,0.14179910553826228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,12288,0.242324431737264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,12288,0.276223103205363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,10240,0.20185421572791207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,8192,0.11452710628509521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,10240,0.23327289687262642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,7168,0.09961422284444173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,8192,0.16316533088684082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,7168,0.14493689272138807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,7168,0.17061600420210096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,6144,0.09182488918304443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,5120,0.07175555494096544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,6144,0.12616977426740858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,6144,0.15007910463545057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,4096,0.05749333567089505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,5120,0.1090000006887648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,5120,0.12898311350080702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,3584,0.051701333787706166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,4096,0.09060711330837673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,4096,0.10859911309348212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,3072,0.04407022065586514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,3584,0.08065244224336413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,3584,0.09837866491741604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,8192,0.1918791135152181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,2560,0.03838933176464505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,3072,0.07176178031497531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,2048,0.02968177861637539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,3072,0.08904621998469035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,2560,0.07830311192406549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,1536,0.02456088860829671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,2560,0.0619680020544264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,2048,0.0672560003068712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,2048,0.050534221861097545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,1024,0.017418666018380057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,1536,0.041506667931874595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,1024,0.03196888830926683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,768,0.014203555054134794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,1024,0.04728977878888448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,1536,0.05865155325995552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,512,0.013175110850069257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,768,0.02735200027624766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,768,0.041724443435668945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,256,0.010209778116808997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,512,0.03769955701298184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,512,0.023752000596788194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,256,0.02100177771515316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,128,0.010055111514197456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,256,0.03180711136923896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,64,0.009464888936943477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,128,0.020264888803164165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,12288,32,0.009352888498041365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,12288,128,0.027722666660944622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,64,0.020394666327370536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,12288,32,0.019920888874265883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,65536,0.7121120029025607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,65536,1.2174524731106229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,51200,0.952365345425076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,65536,1.2966640260484483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,51200,0.5587689081827799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,16384,0.1927173270119561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,16384,0.29372088114420575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,12288,0.14645777808295354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,16384,0.3323129018147786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,12288,0.2166986730363634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,12288,0.2506595452626546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,10240,0.12223822540707058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,51200,1.018612437778049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,10240,0.18016444312201607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,8192,0.09817866484324138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,10240,0.21278756194644502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,8192,0.14578578207227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,7168,0.0866195559501648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,8192,0.17421244250403511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,6144,0.07432355483373006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,7168,0.1291395558251275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,7168,0.1564133299721612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,5120,0.061475555102030434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,6144,0.1131617758009169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,6144,0.1376106606589423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,4096,0.05175288849406772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,5120,0.11865599950154622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,5120,0.09673688809076945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,3584,0.04461422231462267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,4096,0.0805164443122016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,4096,0.10058489110734727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,3072,0.03913777735498216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,3584,0.09100000063578288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,3584,0.07202133205201891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,2560,0.03294310967127482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,3072,0.06353600157631768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,3072,0.08243822389178805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,2048,0.02835822105407715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,2560,0.07261777586407132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,2560,0.05487377776039971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,2048,0.04552711049715678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,1536,0.02152266601721446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,2048,0.06313688887490167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,1024,0.016016888949606154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,1536,0.03706399930848016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,1536,0.055239998632007174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,768,0.013365333278973898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,1024,0.028547555208206177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,1024,0.04417066772778829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,512,0.01057866629627016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,768,0.03901244534386529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,768,0.025260445144441392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,512,0.020964443683624268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,256,0.007993777592976889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,512,0.0339955555068122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,128,0.007941332956155142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,256,0.019300444258583915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,256,0.02959022257063124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,10240,128,0.025972444150182936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,64,0.007638221813572778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,10240,32,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,64,0.018604444132910836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,32,0.01891288823551602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,65536,0.5533315340677897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,10240,128,0.01920355525281694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,65536,1.0605955123901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,51200,0.43052977985805935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,16384,0.1447182231479221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,16384,0.25771111912197536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,51200,0.8332382308112251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,12288,0.10869333479139541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,65536,1.168824937608507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,16384,0.300343116124471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,12288,0.19039200411902535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,10240,0.09565689166386922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,51200,0.9180942111545138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,12288,0.22692622078789604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,8192,0.07752533091439141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,10240,0.15840000576443142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,10240,0.19259022341834175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,7168,0.06543378035227458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,8192,0.12855554951561823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,8192,0.15803644392225477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,7168,0.11439200242360432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,7168,0.14220978154076472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,6144,0.10054310825135972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,5120,0.05047466688685947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,6144,0.1258080005645752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,5120,0.0855413344171312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,4096,0.04001599881384108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,5120,0.10865155855814616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,4096,0.06946400139066908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,6144,0.060524443785349526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,3584,0.03599911265903049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,4096,0.09175111187828912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,3072,0.03151822090148926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,3584,0.06369333134757148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,3584,0.0833031137784322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,2560,0.02755111124780443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,3072,0.07558400101131864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,3072,0.056221332814958364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,2560,0.04821866750717163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,2048,0.023194667365815904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,2560,0.06691910823186238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,1536,0.018970666660202872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,2048,0.040010667509502836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,2048,0.058520886633131236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,1024,0.014056889547242058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,1536,0.03396711084577773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,1536,0.051091555092069835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,768,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,1024,0.04139022363556756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,1024,0.026156443688604567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,768,0.022356443934970435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,512,0.009400889277458191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,768,0.036333332459131874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,256,0.009069333473841349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,512,0.01921066641807556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,512,0.032280001375410296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,256,0.01796000036928389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,128,0.006715555571847492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,256,0.02899822261598375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,64,0.007318221860461765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,128,0.01756266587310367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,8192,128,0.02511022157139248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,8192,32,0.007661333514584436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,64,0.01722755531469981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,8192,32,0.017416889468828838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,65536,0.4384008778466119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,51200,0.34407467312282985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,65536,0.9825475480821398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,51200,0.7754284540812174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,16384,0.12166399425930446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,65536,1.0417235692342122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,16384,0.23955821990966797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,12288,0.09150666660732693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,51200,0.8189893298678927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,16384,0.26884622044033474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,12288,0.17625421947903105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,10240,0.07928711175918579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,12288,0.20359910859002006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,8192,0.06294666396247016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,10240,0.14701067076789007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,10240,0.1720186604393853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,8192,0.14167200194464788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,7168,0.05568355321884155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,7168,0.10652799738778009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,7168,0.12782488928900823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,6144,0.05777600076463488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,6144,0.1132871045006646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,6144,0.09463466538323297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,5120,0.04782399866316053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,5120,0.08102044132020739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,5120,0.09826311137941147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,8192,0.11968799432118733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,4096,0.038635555240843035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,3584,0.03422310948371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,4096,0.06414933337105645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,4096,0.08319466643863253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,3072,0.029861334297392104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,3584,0.07540000147289701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,3584,0.05758577585220337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,3072,0.05209066801600986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,2560,0.025423111187087163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,3072,0.06918844249513414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,2048,0.02139555580086178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,2560,0.04459555612670051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,2560,0.06192800071504381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,1536,0.017139555679427255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,2048,0.03760266635153029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,2048,0.05371910995907254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,1024,0.012435555458068848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,1536,0.04723377691374885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,1536,0.031896889209747314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,768,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,1024,0.024119999673631456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,1024,0.03805333375930786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,512,0.007649777664078607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,768,0.03334844443533156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,768,0.021505778034528095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,512,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,256,0.006301333506902059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,512,0.03126666612095303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,256,0.01719999975628323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,256,0.027817779117160376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,128,0.005582222094138463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,7168,128,0.024696000748210486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,64,0.005219555563396878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,7168,32,0.005323555734422471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,64,0.016529776983790927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,32,0.016597333881590102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,65536,0.4169742266337077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,7168,128,0.01683466633160909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,65536,0.9133680131700305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,51200,0.32656799422370064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,51200,0.7136471006605359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,65536,1.039322641160753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,16384,0.10766400231255426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,16384,0.22182933489481607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,12288,0.08856888612111409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,51200,0.8179413477579752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,12288,0.1641555494732327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,10240,0.07209689087337919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,12288,0.2023822201622857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,10240,0.13735555277930364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,10240,0.1712657743030124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,8192,0.058054222000969775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,16384,0.26778311199612087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,8192,0.1411857737435235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,8192,0.11380178398556179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,7168,0.04693511128425598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,6144,0.04162133402294583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,7168,0.10133688979678684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,7168,0.12755289342668322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,6144,0.08918133046891953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,5120,0.0373982224199507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,6144,0.11288266711764866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,4096,0.031164444155163232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,5120,0.07622933387756348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,5120,0.09781510962380303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,4096,0.060348444514804415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,3584,0.02712000078625149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,4096,0.08240355385674371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,3072,0.023357333408461675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,3584,0.05380444394217598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,3584,0.07566577858395047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,2560,0.020971554848882887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,3072,0.048327998982535474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,3072,0.06842933098475139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,2048,0.017696888910399545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,2560,0.06186311112509834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,2560,0.04311555624008179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,2048,0.03592000073856778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,1536,0.014858666393491956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,2048,0.053393777873780995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,1024,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,1536,0.029861334297392104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,1536,0.04764799939261543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,1024,0.023099554909600153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,1024,0.0369777778784434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,768,0.032927112446890935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,768,0.020271110865804885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,512,0.007689778175618913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,512,0.017779555585649278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,256,0.00673333348499404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,256,0.016572443975342643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,256,0.02719199988577101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,128,0.0064399999876817065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,768,0.008995555341243744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,128,0.016201777590645682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,64,0.006113777971929974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,128,0.024264888630972967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,6144,32,0.006311111152172089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,64,0.01592088904645708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,6144,512,0.03052266769938999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,6144,32,0.015863999724388123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,65536,0.36377867062886554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,51200,0.2812142107221815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,65536,0.8301502333747016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,51200,0.6552426550123427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,16384,0.09665333562427098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,65536,1.03911558787028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,16384,0.20558399624294707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,12288,0.07452888621224298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,51200,0.8174293306138781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,10240,0.0628239976035224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,16384,0.2673315472073025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,12288,0.2018773290846083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,12288,0.15050133069356283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,8192,0.04866577850447761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,10240,0.12631377908918592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,7168,0.04517066809866163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,10240,0.17127554946475557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,8192,0.10406221946080525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,8192,0.1410355567932129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,7168,0.09297421905729507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,7168,0.12751021650102404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,6144,0.0817742215262519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,5120,0.03140177660518222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,6144,0.1125333309173584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,4096,0.026457778281635706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,5120,0.06972977850172254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,5120,0.09726755486594306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,4096,0.055332445436053805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,3584,0.023909333679411147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,3584,0.04902844296561348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,6144,0.036192887359195285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,3072,0.019896000623703003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,3584,0.07458044422997369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,3072,0.04430222180154589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,3072,0.06813600328233507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,2560,0.017539555827776592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,2560,0.038896001047558255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,2560,0.06124355395634969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,2048,0.015480000111791821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,4096,0.0823084447118971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,2048,0.03334488802485996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,1536,0.012513777448071374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,2048,0.05259644322925144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,1024,0.009716444545321995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,1536,0.02695200012789832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,1536,0.04665155543221367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,768,0.008011555506123437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,1024,0.021848888860808477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,1024,0.036335110664367676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,512,0.006981333096822103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,768,0.019223110543357003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,768,0.03290044599109226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,256,0.005951111101441913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,512,0.016896888613700867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,512,0.02993866801261902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,128,0.005236444373925527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,256,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,256,0.02685600022474925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,64,0.005008000052637524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,128,0.015471110741297403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,5120,128,0.0233271105421914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,5120,32,0.004955555415815777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,64,0.015146666102939181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,5120,32,0.014793778459231058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,65536,0.25942667325337726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,51200,0.19516089227464464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,65536,0.7504862149556478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,51200,0.5935475561353896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,16384,0.06797333558400472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,65536,0.910719977484809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,16384,0.18626311090257433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,12288,0.052413334449132286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,16384,0.23479999436272514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,51200,0.7168213526407877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,12288,0.13686488734351263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,10240,0.04714044597413805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,12288,0.1774533324771457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,10240,0.11481066544850667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,10240,0.15137155850728354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,8192,0.09433155589633518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,8192,0.1251386668947008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,7168,0.031607998741997614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,7168,0.08389688862694634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,7168,0.11311644977993435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,6144,0.02940800123744541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,6144,0.07378844420115153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,6144,0.10024266772800022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,5120,0.024350222614076402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,5120,0.06262577904595269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,8192,0.038600001070234514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,5120,0.08663200007544623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,4096,0.02120622164673275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,3584,0.019692444139056735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,4096,0.050075554185443454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,4096,0.07397600015004475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,3072,0.017312000195185345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,3584,0.044681777556737266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,3584,0.06722578075197008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,2560,0.015392000476519266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,3072,0.06199555264578926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,2560,0.03579911258485582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,2560,0.0557875567012363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,2048,0.013087111214796701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,2048,0.029911110798517864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,2048,0.04796533452139961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,1536,0.010835555692513784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,1536,0.02570933269129859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,1536,0.041060444381501936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,1024,0.008296888735559251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,1024,0.03384000062942505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,1024,0.019913777709007263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,768,0.007063999772071838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,3072,0.04038133223851522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,768,0.03081688947147793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,512,0.006247111078765657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,512,0.015885333220163982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,256,0.005220444252093633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,256,0.014820444915029736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,256,0.02576622201336755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,128,0.004590222405062782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,128,0.014434667097197639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,768,0.01761777698993683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,128,0.023035556077957153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,64,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,4096,32,0.004566222015354368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,64,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,4096,512,0.028622223271263972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,65536,0.27021511395772296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,4096,32,0.014856000741322836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,51200,0.2220311164855957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,65536,0.7125262154473199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,51200,0.5632524490356445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,65536,0.9107679790920682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,16384,0.07219733132256402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,51200,0.7169991069369845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,16384,0.1759751107957628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,12288,0.05810311105516222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,16384,0.23571022351582846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,10240,0.04838133189413282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,12288,0.12949955463409424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,12288,0.17744889524247912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,8192,0.040059556563695274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,10240,0.15102488464779323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,8192,0.08964711427688599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,7168,0.035262223747041486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,8192,0.12565688292185465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,7168,0.07973421944512261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,7168,0.11280532677968343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,6144,0.030672887961069744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,6144,0.0696835584110684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,6144,0.1000231107076009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,5120,0.026251556144820318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,5120,0.05847733550601535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,5120,0.08665511343214248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,10240,0.10911999808417426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,4096,0.021923555268181696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,3584,0.02008888953261905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,4096,0.047608888811535306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,4096,0.07424177726109822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,3072,0.018936889039145578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,3584,0.04250133368704054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,3584,0.06759911113315158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,2560,0.01640533407529195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,3072,0.06141510936948988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,2560,0.03421955638461643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,2048,0.01354311075475481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,2560,0.055458668205473155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,3072,0.0395706660217709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,1536,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,2048,0.028466665082507666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,2048,0.04771022333039177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,1024,0.008721777962313758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,1536,0.04225777917438083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,1536,0.02402399977048238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,1024,0.018989332848125033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,768,0.007344889144102733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,1024,0.03380799955791897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,768,0.017208000024159748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,512,0.006069333189063602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,512,0.01518133282661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,768,0.03131555517514547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,256,0.005234666582610872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,512,0.02811466654141744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,128,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,256,0.025754666990704004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,256,0.014851555228233337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,64,0.004282666577233208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,128,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3584,32,0.0042257776690853965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3584,128,0.023048889305856492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,64,0.014209777116775513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3584,32,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,65536,0.19008089436425102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,51200,0.1496071153216892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,65536,0.6973982387118869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,51200,0.5444969071282281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,16384,0.05377511183420817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,65536,0.9106818305121528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,16384,0.17141600449879965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,12288,0.04233333468437195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,16384,0.23507022857666016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,12288,0.12696444988250732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,51200,0.7175208727518717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,10240,0.035354667239718966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,12288,0.1781031158235338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,8192,0.030254221624798242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,10240,0.15063644780053034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,10240,0.10729510916603936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,7168,0.02645955483118693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,8192,0.08782044384214614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,8192,0.1250657770368788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,6144,0.023856000767813787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,7168,0.0788008901807997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,7168,0.11249066723717584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,5120,0.021075555019908484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,6144,0.06846221950319079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,6144,0.10037599669562446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,5120,0.0575431121720208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,4096,0.017456000049908955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,5120,0.08623377482096355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,3584,0.015999111864301894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,4096,0.04641866683959961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,4096,0.07403110795550875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,3584,0.04186577929390801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,3072,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,3584,0.06702133019765218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,2560,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,3072,0.03748800026045905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,3072,0.06213421954049004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,2560,0.0332897769080268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,2048,0.01094044413831499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,2560,0.055833776791890465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,2048,0.027451554934183758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,1536,0.009152000149091085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,2048,0.0470071103837755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,1024,0.007293333609898885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,1536,0.024013333850436743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,1536,0.04251377781232198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,1024,0.03360533383157518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,768,0.0064275554484791225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,768,0.01687733332316081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,1024,0.019075555933846366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,512,0.005582222094138463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,512,0.0281368891398112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,256,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,512,0.015343111422326831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,256,0.014127999544143677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,128,0.004869333157936732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,256,0.02590488890806834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,128,0.013815110756291283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,64,0.00461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,128,0.022755554980701868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,3072,32,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,64,0.01348533315791024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,3072,768,0.030657778183619182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,65536,0.1593902243508233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,3072,32,0.013802666631009845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,51200,0.14706933498382568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,65536,0.6567173533969456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,51200,0.5153617858886719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,16384,0.046521776252322726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,65536,0.9096942477756076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,51200,0.716543992360433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,16384,0.16395555602179632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,16384,0.23496445020039877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,12288,0.037864890363481306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,10240,0.0317795541551378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,12288,0.177565336227417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,10240,0.10157955355114406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,8192,0.025389333566029865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,10240,0.1507342259089152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,8192,0.08311111397213407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,7168,0.022496889034907024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,8192,0.12520266903771296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,7168,0.0742595526907179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,6144,0.019523556033770244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,6144,0.06506222486495972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,12288,0.12109955151875813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,5120,0.017649778061442904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,6144,0.09955733352237278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,5120,0.05417955583996243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,5120,0.08615289131800334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,4096,0.04423911041683621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,7168,0.1118897729449802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,3584,0.014002665877342224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,4096,0.07342311408784655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,3584,0.039840890301598444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,3072,0.012532444463835822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,3584,0.0670142239994473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,3072,0.035536001125971474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,3072,0.06150222486919827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,2560,0.011336888704035016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,4096,0.014850666125615438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,2048,0.009304000271691216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,2560,0.03199377655982971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,2560,0.054972443315717906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,1536,0.007872000336647034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,2048,0.04559822214974297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,1536,0.02239111065864563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,1024,0.0062773335311147906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,1536,0.04047733214166429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,1024,0.01797155539194743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,768,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,1024,0.03356533249219259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,768,0.016176889340082806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,512,0.004885333279768626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,512,0.014463111758232117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,512,0.027833776341544256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,256,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,2048,0.026471111509535048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,256,0.013545778062608508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,128,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,256,0.025095999240875244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,128,0.013150222599506378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,64,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,768,0.03018222252527873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2560,32,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2560,128,0.022760889596409265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,64,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,65536,0.15377510918511286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2560,32,0.013165333204799227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,51200,0.11973955896165635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,65536,0.6132071283128526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,16384,0.04288533329963684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,51200,0.48162841796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,51200,0.6150266859266493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,65536,0.7820231119791666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,16384,0.15398666593763563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,12288,0.03310133351220025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,16384,0.2027439938651191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,10240,0.027253333065244887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,12288,0.11312799983554417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,12288,0.15310311317443848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,8192,0.02233244478702545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,10240,0.09495555692248875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,10240,0.13016888830396864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,7168,0.020181333025296528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,8192,0.07815555731455485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,7168,0.06936266687181261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,7168,0.09767644272910224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,6144,0.017719111508793302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,6144,0.08657777971691555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,5120,0.01552888916598426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,5120,0.04997244477272034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,5120,0.07529689206017388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,4096,0.013214222258991666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,8192,0.1081119974454244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,4096,0.04142399960094028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,3584,0.012150222228633033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,3584,0.037131554550594754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,6144,0.06039110819498698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,3072,0.010339555641015371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,3584,0.05925866630342272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,3072,0.03338311115900675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,2560,0.00925155563486947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,3072,0.05483022332191467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,2560,0.029874665869606867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,2560,0.047395557165145874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,2048,0.007672000262472365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,2048,0.025306666890780132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,2048,0.04076977901988559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,4096,0.06537333461973402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,1536,0.007153777612580194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,1024,0.0058133333093590206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,1536,0.021532444490326777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,1536,0.03644711110326979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,768,0.0052791109515561005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,1024,0.017247110605239868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,1024,0.030538668235143025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,768,0.015575110912322998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,768,0.027804444233576458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,512,0.02605955633852217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,256,0.004210666649871402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,256,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,256,0.02439466615517934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,128,0.003843555433882607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,128,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,2048,128,0.021632888250880774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,512,0.004907555464241239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,64,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,2048,32,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,64,0.012799999780125089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,512,0.014097778333557976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,2048,32,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,65536,0.13030132982465956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,51200,0.10239822334713405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,65536,0.5742488967047797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,51200,0.4521928893195258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,16384,0.03549777799182468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,65536,0.778257793850369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,16384,0.14473866091834173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,12288,0.026317333181699116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,16384,0.2017608880996704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,51200,0.6153093443976508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,10240,0.023039999935362074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,12288,0.10621955659654404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,12288,0.15274843904707167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,10240,0.0892524454328749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,8192,0.01878666712178124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,10240,0.12968711058298746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,7168,0.016599110431141324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,8192,0.10742044448852539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,8192,0.07279111279381646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,6144,0.014717333846622042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,7168,0.0650213360786438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,7168,0.09667199850082397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,5120,0.013003555436929068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,6144,0.05552444524235196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,6144,0.08627822001775105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,4096,0.01127466642194324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,5120,0.0467582212554084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,4096,0.03869333200984531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,4096,0.064264886909061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,3584,0.010047111246320937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,3584,0.03458666801452637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,3584,0.05777244435416328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,3072,0.009199110998047723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,3072,0.0316880014207628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,3072,0.05379199981689453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,2560,0.008298666940795051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,2560,0.02793688906563653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,2048,0.0075253330998950535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,2560,0.04811733298831516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,1536,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,2048,0.023697778582572937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,5120,0.07484889030456543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,1024,0.005592000153329637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,1536,0.020426667398876615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,2048,0.04090044564670987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,1536,0.03604266709751553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,768,0.004921777794758479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,1024,0.016531555189026725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,1024,0.03092355529467265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,768,0.014835556348164877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,512,0.004646222210592694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,768,0.028155555327733357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,256,0.004231111043029361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,512,0.013454222016864352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,512,0.026456000076399908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,256,0.013162666724787818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,256,0.02423022190729777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,128,0.012764444781674279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1536,128,0.021735999319288466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,64,0.003899555653333664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,32,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,64,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1536,32,0.012406222522258759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,65536,0.08449333243899876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,65536,0.5337653160095215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1536,128,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,51200,0.06894844770431519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,65536,0.7680951224433051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,16384,0.030191110240088567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,51200,0.4194071027967665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,16384,0.1346773306528727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,51200,0.6052791277567545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,16384,0.19942755169338652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,12288,0.0987573332256741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,10240,0.019428445233239066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,12288,0.15079555246565077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,10240,0.08291288879182604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,8192,0.016590222716331482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,10240,0.12823822763231066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,8192,0.06776089138454862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,7168,0.014847111370828418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,12288,0.02309422194957733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,8192,0.1063199970457289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,6144,0.013533333109484779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,7168,0.05964177846908569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,7168,0.0950346655315823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,5120,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,6144,0.0848239991399977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,6144,0.051735109753078885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,5120,0.0429075558980306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,4096,0.010072000324726105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,5120,0.07391733593410917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,3584,0.00870400004916721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,4096,0.035396443472968206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,3584,0.03197244471973843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,3584,0.057803557978736035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,3072,0.008032889001899296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,3072,0.029491557015313044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,3072,0.052038222551345825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,2560,0.007461333440409766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,2560,0.02613688839806451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,2560,0.045927110645506114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,2048,0.0063902222447925145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,2048,0.022361778550677832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,2048,0.040227555566363864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,1536,0.00592533333433999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,1536,0.01928622192806668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,1024,0.004954666727119022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,4096,0.06393955813513862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,1024,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,768,0.004586666822433472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,1024,0.030336889955732558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,768,0.01458755632241567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,768,0.028559999333487615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,512,0.003962666624122196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,512,0.025751110580232408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,512,0.013439999686347114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,256,0.003934222377008862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,1536,0.035980443159739174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,256,0.023786667320463393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,256,0.012797333300113678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,128,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,64,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,128,0.012464000119103326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,1024,128,0.021407999926143225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,1024,32,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,64,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,65536,0.07915644513236152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,1024,32,0.011970666547616323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,51200,0.06287911203172472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,65536,0.5299217965867784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,51200,0.41657598813374835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,65536,0.7678435643513998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,16384,0.024939555260870192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,16384,0.13384089205000135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,51200,0.6045911047193739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,12288,0.01759288873937395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,16384,0.199109329117669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,10240,0.015856888559129503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,12288,0.09756622049543594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,12288,0.15025688542260066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,8192,0.013267555170589022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,10240,0.08196889029608832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,10240,0.12750755416022405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,8192,0.06683733065923055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,7168,0.011830222275522021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,8192,0.10522133111953735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,7168,0.05901155869166056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,6144,0.010685332947307162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,7168,0.09481155872344971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,5120,0.009263111485375298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,6144,0.05086311035686069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,5120,0.04231111208597819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,6144,0.08526755703820123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,4096,0.008007110820876228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,5120,0.07390755414962769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,3584,0.007644444704055786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,4096,0.06263377931382921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,3584,0.03188800149493747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,3584,0.05579288800557455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,3072,0.006928000185224745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,3072,0.028455999162462022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,3072,0.05136533247099983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,2560,0.006251555350091722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,2560,0.025751999682850305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,2560,0.04654666781425476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,2048,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,2048,0.021951110826598272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,2048,0.03975911272896661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,1536,0.005232000102599462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,4096,0.03465155429310269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,1536,0.03565244542227851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,1024,0.004514666481150521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,1024,0.015549333559142219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,1024,0.029808001385794744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,768,0.004295111116435793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,768,0.027423999375767175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,768,0.014135110709402295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,512,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,512,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,512,0.02574311031235589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,256,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,256,0.01238933371173011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,256,0.023737778266270954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,128,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,1536,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,768,128,0.021301334102948506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,128,0.012071111136012606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,64,0.003177777760558658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,768,32,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,65536,0.051114665137396924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,64,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,768,32,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,51200,0.04262400004598829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,65536,0.49186309178670246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,51200,0.38691287570529515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,16384,0.018608000543382432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,65536,0.7674061987135145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,16384,0.12431911627451579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,12288,0.017200888858901132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,51200,0.6056035359700521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,16384,0.19864355193244088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,10240,0.014873777826627096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,12288,0.09034666750166152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,12288,0.15026666058434382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,10240,0.07573066817389594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,8192,0.012658666405412885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,10240,0.12761955791049534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,7168,0.011367999845080905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,8192,0.0616204473707411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,8192,0.10578044255574544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,6144,0.010041777458455827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,7168,0.054197334580951266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,7168,0.09486755397584702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,5120,0.008975110948085785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,6144,0.04589244392183092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,6144,0.08456355333328247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,4096,0.007673777639865875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,5120,0.03943199912707011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,5120,0.0734142197502984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,3584,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,4096,0.03251377741495768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,4096,0.0625422199567159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,3072,0.006613333192136552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,3584,0.02963199880388048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,3584,0.05602755811479357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,2560,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,3072,0.027071999178992376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,3072,0.05100889007250468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,2048,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,2560,0.024375110864639282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,2560,0.045583999819225736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,1536,0.004866666677925321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,2048,0.020959999826219346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,2048,0.03945599993069967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,1024,0.004209777961174647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,1536,0.03529955612288581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,1024,0.01514400045077006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,1024,0.029551111989551123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,768,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,768,0.01386844449573093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,768,0.027450667487250432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,512,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,512,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,512,0.025386666258176167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,256,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,1536,0.018200000127156574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,256,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,128,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,256,0.0233271105421914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,128,0.011815999945004782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,512,128,0.02106133268939124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,64,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,512,32,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,65536,0.03659644391801622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,64,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,512,32,0.011487111449241638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,51200,0.03019555409749349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,65536,0.4907742076449924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,65536,0.7670293384128146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,51200,0.38682932323879665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,16384,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,16384,0.1989377737045288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,51200,0.6046880086263021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,12288,0.013233777549531726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,16384,0.12355021635691325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,10240,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,12288,0.14959200223286948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,10240,0.0753199987941318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,10240,0.12794221772087946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,8192,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,8192,0.0612000028292338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,7168,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,8192,0.10544088813993667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,7168,0.05312888820966085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,6144,0.009783999787436591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,7168,0.09524533483717178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,12288,0.08978844351238674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,5120,0.00870488915178511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,6144,0.045293334457609385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,6144,0.08496799733903672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,5120,0.0377768874168396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,4096,0.007637333538797166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,5120,0.0725875563091702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,3584,0.0070017774899800616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,4096,0.03190133306715224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,4096,0.062055998378329806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,3072,0.006598222172922558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,3584,0.029118223322762385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,3584,0.05483999848365784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,2560,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,3072,0.05071555574735006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,3072,0.027531555957264368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,2048,0.0053288886944452924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,2560,0.023815110325813293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,2560,0.04562044474813673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,1536,0.004566222015354368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,2048,0.039192001024881996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,2048,0.021002666817771062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,1024,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,1536,0.01819022165404426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,1536,0.03530755639076233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,768,0.003847111016511917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,1024,0.030012445317374334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,1024,0.01520088811715444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,768,0.013518222504191928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,512,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,768,0.02711733347839779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,256,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,512,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,512,0.025489777326583862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,128,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,256,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,128,0.020983111527231004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,128,0.01148622234662374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,64,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,256,32,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,64,0.01146133326821857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,256,32,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,65536,0.029696888393825952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,65536,0.49090576171875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,256,256,0.023034666975339253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,51200,0.025012445118692186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,51200,0.38600177235073513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,65536,0.7688150935702853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,16384,0.010089777410030365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,51200,0.6047351095411513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,12288,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,16384,0.12302933798895942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,16384,0.1983635558022393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,10240,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,12288,0.08942311339908177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,12288,0.1501973337597317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,10240,0.07503733370039198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,10240,0.12726133399539524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,8192,0.009741333623727163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,7168,0.00905600024594201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,8192,0.06035111347834269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,8192,0.10494222243626912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,6144,0.008798222574922774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,7168,0.05312977896796333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,7168,0.0945742196506924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,5120,0.007661333514584436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,6144,0.04547377758555942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,6144,0.08452710840437148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,5120,0.03896977835231357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,4096,0.007663110891977946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,5120,0.07223733266194661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,4096,0.032624890406926475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,3584,0.0068604445291890045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,4096,0.061030222309960254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,3584,0.029278222057554457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,3072,0.006275555739800136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,3072,0.026512889398468867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,3584,0.055198222398757935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,3072,0.05212355653444926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,2560,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,2048,0.0052871112194326185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,2560,0.024122666981485155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,2560,0.045904888047112354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,2048,0.02068177858988444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,1536,0.004914666629499859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,2048,0.039452443520228066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,1024,0.004307555655638377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,1536,0.018260444204012554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,1536,0.03570311268170675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,768,0.0038533334930737815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,1024,0.014867555763986377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,1024,0.029636445972654555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,512,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,768,0.027101332942644756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,768,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,512,0.012458666331238218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,256,0.003169777699642711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,512,0.024999999337726172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,128,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,256,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,256,0.023080888721677992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,128,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,1024,128,128,0.020650666620996263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,128,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,65536,0.026701332794295415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,32,0.011544889046086205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,128,64,0.011704888608720569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,51200,0.022301332818137273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,16384,0.009408888717492422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,65536,0.4900053342183431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,16384,0.12310755252838135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,12288,0.008373333348168267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,51200,0.386294232474433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,10240,0.00868800034125646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,8192,0.008018666671382057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,12288,0.08950310945510864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,7168,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,8192,0.060263997978634305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,10240,0.07472444242901273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,6144,0.007326222128338284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,5120,0.007363555332024892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,7168,0.051627556482950844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,4096,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,6144,0.04526488979657491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,5120,0.03723288906945123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,3584,0.006667555620272954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,4096,0.03195555673705207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,3584,0.028833776712417603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,2560,0.005614222337802251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,3072,0.026199999782774184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,2048,0.005141333159473208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,2560,0.023704888092146978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,1536,0.004600000050332812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,2048,0.020915556285116408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,1024,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,1536,0.017883555756674874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,768,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,1024,0.014819555812411837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,512,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,768,0.013546666337384118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,256,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,512,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,3072,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,128,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,64,0.0025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,256,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,64,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,64,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,128,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,65536,0.025980444418059453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,64,32,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,51200,0.018239110708236694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,16384,0.011115555961926779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,16384,0.12287378311157227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,51200,0.3860062228308783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,12288,0.009692444569534725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,65536,0.4907671080695258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,10240,0.008988444175985124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,8192,0.008852444589138031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,12288,0.0885759989420573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,7168,0.008262222011884054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,10240,0.07467199696434869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,8192,0.06083377864625719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,6144,0.007829333345095316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,5120,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,7168,0.05225511060820686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,6144,0.04512888855404324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,4096,0.007358222372002072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,5120,0.037074665228525795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,3584,0.006966222491529252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,4096,0.03161244591077169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,3072,0.006248888870080312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,3584,0.028889778587553237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,2560,0.005607111172543631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,3072,0.026390223039521113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,2048,0.005575110928879843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,2560,0.023751111494170293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,1536,0.004944889081848992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,2048,0.0207013338804245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,1536,0.017895110779338412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,768,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,1024,0.014827556080288358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,512,0.003547555456558863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,768,0.013510222236315409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,512,0.012516444755925072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,128,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,256,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,64,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,128,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1024,32,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,64,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1024,32,32,0.011136000355084738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,16384,0.8327777650621203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,51200,2.5093903011745877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,12288,0.6188213560316298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,16384,1.1019751230875652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,51200,3.745698716905382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,12288,0.8129555384318033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,10240,0.5204213460286459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,8192,0.4035608768463135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,10240,0.648231135474311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,8192,0.5256142086452907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,7168,0.35799111260308164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,6144,0.3273679945203993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,7168,0.45944446987575954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,6144,0.4055013391706679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,5120,0.24587644471062553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,5120,0.33040444056193036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,4096,0.1984808842341105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,3584,0.17599199877844918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,4096,0.2808764510684543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,3584,0.2341386742062039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,3072,0.15293778313530818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,2560,0.1291831069522434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,3072,0.20156444443596733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,2048,0.10643111334906684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,2048,0.14343733257717556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,1536,0.11340088314480251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,1536,0.0831893351342943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,1024,0.062330669826931424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,768,0.04940711127387153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,1024,0.08722044361962213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,512,0.03717422154214647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,768,0.0718640022807651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,256,0.028607110182444256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,512,0.06295200188954671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,128,0.025527111358112756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,256,0.05797421932220459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,64,0.021722666091389124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,2560,0.17422489325205484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,128,0.05459288756052653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,65536,32,0.022460444105996027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,64,0.05378489030732048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,65536,32,0.05373600125312805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,51200,2.0944186316596136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,65536,2.696702321370443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,16384,0.6870489120483398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,65536,3.6350390116373696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,12288,0.5154568884107802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,51200,2.991419474283854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,16384,0.8604426913791232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,10240,0.5195866690741645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,8192,0.41736711396111387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,12288,0.6338915295071071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,10240,0.5168835322062174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,7168,0.36352178785536027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,6144,0.3085430992974175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,7168,0.3654346730973985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,8192,0.41479023297627765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,5120,0.25062934557596844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,4096,0.20342489083607992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,5120,0.2652711073557536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,3584,0.17930932839711508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,3072,0.15425777435302734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,4096,0.21546222103966606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,2560,0.1304373343785604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,3584,0.18905866146087646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,3072,0.16285510857899985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,2048,0.10872444179322983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,1536,0.0849271085527208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,2560,0.14155822330051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,2048,0.1172222163942125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,1024,0.06143289142184787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,6144,0.3158666557735867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,1536,0.09231022331449722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,768,0.04959644542800056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,1024,0.07186666462156507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,512,0.0365555551317003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,256,0.023039999935362074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,768,0.05963999695248074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,128,0.024538666009902954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,512,0.05236977669927809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,64,0.021355556117163763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,256,0.047988444566726685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,51200,32,0.022328888376553852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,128,0.045147554741965405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,64,0.044290665123197764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,51200,32,0.04411111275355021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,65536,0.9091315799289279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,51200,0.699187543657091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,16384,0.23039287990993926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,65536,1.3490364286634657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,12288,0.17453155252668592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,16384,0.3187546730041504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,10240,0.172888888253106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,51200,1.0727760526869032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,12288,0.23926133579678008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,8192,0.13621777958340114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,10240,0.202930662367079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,7168,0.11474400096469456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,8192,0.162352893087599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,6144,0.10262044270833333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,5120,0.08791644705666436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,7168,0.14446489016215006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,4096,0.07163821988635592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,6144,0.12557689348856607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,3584,0.06337155236138238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,4096,0.08841955661773682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,3072,0.0517119997077518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,3584,0.07930666870541044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,2560,0.04519644379615784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,3072,0.06914755370881823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,2048,0.0351493325498369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,2560,0.06053421894709269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,1536,0.02860355708334181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,2048,0.05059022373623318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,1024,0.019853333632151287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,1536,0.0404097768995497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,768,0.018806222412321303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,1024,0.03218577636612786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,512,0.01239466667175293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,768,0.02751111156410641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,256,0.009711999860074785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,512,0.024064888556798298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,256,0.02162933349609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,128,0.008302222523424361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,64,0.006994666324721442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,128,0.020928000410397846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,16384,32,0.007664889097213745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,32,0.020270221763186984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,5120,0.10711110962761773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,65536,0.6610950893825954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,16384,64,0.020928889513015747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,51200,0.5164657698737251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,65536,1.1031306584676106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,16384,0.17344178093804252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,12288,0.13373155064053005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,16384,0.2643013265397814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,51200,0.8646995756361219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,12288,0.19612088468339708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,10240,0.11131200525495742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,8192,0.09145244624879624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,10240,0.16399199432796902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,7168,0.08048888709810045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,8192,0.13353421952989367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,6144,0.07109422153896756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,5120,0.060896886719597705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,7168,0.1184320052464803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,4096,0.04937600096066793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,6144,0.10335289107428657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,3584,0.044237332211600415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,5120,0.08919911252127753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,3072,0.03847466574774848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,4096,0.07412711117002699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,2560,0.032638221979141235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,3072,0.055813332398732506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,2048,0.026753778258959453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,2560,0.05037422312630547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,1536,0.02124711043304867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,2048,0.04242488741874695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,1024,0.01607377827167511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,1536,0.0340008901225196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,768,0.013416888813177744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,1024,0.0273324449857076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,512,0.011047999891969891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,768,0.024142222272025213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,256,0.008309333688682979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,512,0.0206657780541314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,128,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,256,0.019211555520693462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,64,0.007039111521508958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,128,0.018564444449212816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,12288,32,0.008303999900817871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,64,0.018562666243977018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,32,0.01831466621822781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,65536,0.5570897526211208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,12288,3584,0.06657244761784871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,51200,0.436539543999566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,16384,0.14680443869696722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,51200,0.7687422434488932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,65536,0.9655875100029839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,16384,0.2306248876783583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,12288,0.1125333309173584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,10240,0.08736711078219944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,8192,0.07154755459891425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,10240,0.14645866552988687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,7168,0.08593244685067071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,8192,0.1187395519680447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,6144,0.07386933432685004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,7168,0.10589777761035496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,5120,0.06287644306818645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,6144,0.09331999884711371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,4096,0.04983911249372694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,5120,0.07876622014575534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,3584,0.043375998735427856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,12288,0.17416711648305258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,4096,0.06655644708209567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,3584,0.058508442507849805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,3072,0.037147555086347796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,2560,0.031982223192850746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,2048,0.02605955633852217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,3072,0.051093333297305636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,1536,0.01994400057527754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,2560,0.04420266548792521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,2048,0.038140445947647095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,1024,0.014862222803963555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,1536,0.030870222383075293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,1024,0.025111999776628282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,512,0.009729777773221334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,768,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,256,0.007104888558387756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,512,0.01927199959754944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,128,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,256,0.018242667118708294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,64,0.005670222143332164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,128,0.01755555636352963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,32,0.005988444305128521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,64,0.017515555024147034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,10240,32,0.016882666283183627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,10240,768,0.012362666428089142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,65536,0.44504891501532656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,51200,0.350125339296129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,16384,0.1167724397447374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,65536,0.8486622174580892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,12288,0.08901777532365586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,51200,0.6656684345669216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,10240,0.07532444265153673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,16384,0.2014035516315036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,12288,0.15191733837127686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,8192,0.06234755780961779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,10240,0.12786044014824763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,7168,0.05542311072349548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,8192,0.10492710934744941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,6144,0.04900000161594815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,5120,0.041486223538716636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,7168,0.0935093363126119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,4096,0.03410311208830939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,6144,0.08177955283059014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,3584,0.030461334519916113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,5120,0.06919999917348225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,3072,0.02716800073782603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,3584,0.05144800080193413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,2560,0.023381332556406658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,3072,0.04551288816663954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,2560,0.03907022211286757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,2048,0.024494222468800012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,1536,0.019313777486483257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,2048,0.033999111917283796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,1024,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,1536,0.028183112541834515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,768,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,1024,0.022317333353890315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,512,0.008709333009190029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,4096,0.05667822228537666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,768,0.0203493336836497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,256,0.006585777633719974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,128,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,256,0.016924444172117446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,128,0.016230222251680162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,64,0.00572622236278322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,8192,32,0.00600355573826366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,64,0.016240888171725802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,32,0.01614755557643043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,65536,0.34628889295789933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,8192,512,0.01753155555990007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,51200,0.27452090051439076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,16384,0.09451022413041855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,65536,0.7724577585856119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,16384,0.1873555580774943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,12288,0.07464622126685248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,51200,0.6185057957967123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,10240,0.07801333400938246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,8192,0.0629955530166626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,12288,0.14242133829328749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,10240,0.11971644560496013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,7168,0.055504000849194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,6144,0.04847200049294365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,7168,0.08832977877722846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,8192,0.09851288795471191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,5120,0.041013333532545306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,6144,0.07726399766074286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,4096,0.03348355491956075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,3584,0.029478222131729126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,5120,0.06429777542750041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,3072,0.02546399997340308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,4096,0.052758223480648465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,2560,0.02127733329931895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,3584,0.04753155509630839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,2048,0.018253333038753934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,3072,0.042167110575569995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,1536,0.014774221513006421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,2560,0.037351999017927386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,1024,0.011352000137170156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,2048,0.03223111232121786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,768,0.009167999857001835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,1536,0.026504000027974445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,512,0.007111999723646376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,1024,0.021376889612939622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,256,0.005640000104904175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,768,0.01922044489118788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,128,0.0052488889131281115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,512,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,64,0.004859555512666702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,256,0.016157334049542744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,7168,32,0.004951111144489712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,128,0.015501333607567681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,64,0.01555022266176012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,7168,32,0.015476443701320224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,65536,0.3388035562303331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,51200,0.26220443513658315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,16384,0.0886204441388448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,65536,0.715686215294732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,51200,0.5643555323282877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,16384,0.1734000047047933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,12288,0.06884266932805379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,10240,0.058303998576270215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,8192,0.04795911245875888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,12288,0.13228621747758654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,10240,0.11230399873521592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,7168,0.042153779003355235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,6144,0.037031110790040754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,7168,0.08273866441514757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,5120,0.03137422270245022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,6144,0.07260355684492323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,4096,0.02509244448608822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,5120,0.06024622254901462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,3584,0.022999111149046157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,4096,0.049963557057910495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,3072,0.020478222105238173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,3584,0.04506577716933357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,2560,0.017782222893502977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,3072,0.039679110050201416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,2048,0.015538665983412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,2560,0.03540622194608053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,8192,0.09289422300126816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,1536,0.01238844460911221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,2048,0.030897776285807293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,1024,0.009872000250551436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,768,0.008030222521887885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,1024,0.020260444945759244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,512,0.0069671107663048645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,768,0.01826844447188907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,256,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,512,0.01648977730009291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,128,0.004898666507667965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,256,0.015525332755512662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,64,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,128,0.015286222100257874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,6144,32,0.004979555391603046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,64,0.014789332946141561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,32,0.014537778165605334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,65536,0.31928443908691406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,6144,1536,0.024990222520298425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,51200,0.24935823016696504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,16384,0.08443466822306316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,65536,0.6561155319213867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,12288,0.06458755334218343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,51200,0.5131973160637749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,16384,0.15743288728925917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,10240,0.05514488948716057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,8192,0.04486844605869717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,12288,0.11942488617367214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,7168,0.04025333457522922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,8192,0.08356710937288071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,6144,0.034966223769717746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,7168,0.0745848880873786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,5120,0.029322667254341975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,6144,0.06510311365127563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,4096,0.023864888482623633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,5120,0.053299556175867714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,3584,0.021299555897712708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,4096,0.045370668172836304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,3072,0.01888177792231242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,3584,0.04100088940726386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,3072,0.03605155481232537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,10240,0.10145866870880127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,2560,0.016524444023768108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,2048,0.013991999957296582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,2560,0.032300445768568255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,1536,0.01184088902340995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,2048,0.028190222051408555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,1024,0.00928177767329746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,1536,0.02253955602645874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,768,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,1024,0.018909333480728995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,512,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,768,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,256,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,512,0.015875554747051664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,128,0.0047484445903036325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,256,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,64,0.004229333251714706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,128,0.014492443866199918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,5120,32,0.004564444637960858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,64,0.014171555638313293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,5120,32,0.013842666314707862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,65536,0.2494257820977105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,51200,0.1950871149698893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,65536,0.5955395698547363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,16384,0.06695110930336846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,51200,0.46678664949205184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,16384,0.14451200432247585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,12288,0.05221244361665514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,10240,0.04447822107209099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,12288,0.11071377992630005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,8192,0.036116444402270846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,8192,0.07753688759273954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,7168,0.03165510959095425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,7168,0.06886577606201172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,6144,0.027649778458807204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,6144,0.059248891141679555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,5120,0.023247111174795363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,10240,0.09384355280134414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,4096,0.020041777027977836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,5120,0.04962755574120415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,3584,0.017942221628295053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,4096,0.04200622108247545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,3072,0.016128000285890367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,3584,0.038485334979163274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,2560,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,3072,0.033668445216284856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,2048,0.0119982221060329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,2560,0.03022222386466132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,1536,0.009640889035330879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,2048,0.026542221506436665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,1024,0.00871022211180793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,1536,0.021406221720907424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,768,0.006996444529957241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,1024,0.017940445078743827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,512,0.006001777946949005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,768,0.0165928883685006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,256,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,512,0.015119110544522604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,128,0.004602666530344221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,256,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,64,0.004253333227501975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,128,0.013707555830478668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,4096,32,0.004548444516128964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,64,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,4096,32,0.013146667016877068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,65536,0.22717422909206816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,51200,0.17971732881334093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,65536,0.5586391025119358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,16384,0.06068266762627495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,51200,0.4410853385925293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,12288,0.04727111260096232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,16384,0.13581155406104192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,10240,0.039889779355790876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,12288,0.10354577832751805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,8192,0.03337600164943271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,10240,0.08799999952316284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,7168,0.028917332490285236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,8192,0.07267644670274523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,6144,0.024774221910370722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,7168,0.06409333149592082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,5120,0.02128177715672387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,6144,0.055677334467569985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,4096,0.01735466718673706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,5120,0.0470124449994829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,4096,0.03956444395913018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,3584,0.016376000311639573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,3072,0.014415999253590902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,3584,0.036341332727008395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,2560,0.012409778104888068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,3072,0.032067557175954185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,2048,0.010800888968838586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,2560,0.028491556644439697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,1536,0.008991111483838823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,1536,0.020443555381562974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,1024,0.006826666494210561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,1024,0.01719466679626041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,768,0.006144888699054718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,768,0.01592177814907498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,2048,0.024688889582951862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,512,0.00535022219022115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,256,0.00444088876247406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,512,0.014491554763582019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,128,0.003929777691761653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,128,0.013543999857372709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,256,0.014404444230927361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,64,0.0036195553839206696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3584,32,0.003896000070704354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,64,0.013528889252079858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3584,32,0.012818666795889536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,65536,0.1887271139356825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,51200,0.1428284380171034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,65536,0.5448195669386122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,16384,0.049421333604388766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,51200,0.4278986718919542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,12288,0.03783911135461595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,16384,0.13356178336673313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,10240,0.03767733441458808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,8192,0.03113244308365716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,10240,0.08638133605321248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,7168,0.02747111188040839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,12288,0.10295822223027547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,8192,0.07146755854288737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,6144,0.024075556132528517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,5120,0.020928000410397846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,7168,0.06272355715433757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,4096,0.017862222260899015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,6144,0.05441866649521721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,3584,0.01611555616060893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,5120,0.0464933349026574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,4096,0.03870044483078851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,3072,0.01440622243616316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,3584,0.03598399957021078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,3072,0.03135644396146139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,2048,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,2560,0.027859555350409612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,1536,0.009150221943855286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,2048,0.02427022159099579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,1024,0.007335999773608313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,1536,0.020404444800482858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,768,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,1024,0.016903999778959487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,512,0.005575110928879843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,768,0.015881778465376962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,256,0.004922666483455234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,512,0.01423288881778717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,128,0.00425155543618732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,2560,0.012898666991127862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,256,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,64,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,128,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,3072,32,0.004168888760937585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,64,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,3072,32,0.012479999827014076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,65536,0.16288267241583929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,51200,0.12675644291771784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,16384,0.04424533247947693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,65536,0.5053786701626247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,16384,0.12408799595303005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,12288,0.034318221939934626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,51200,0.39812800619337296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,10240,0.028918223248587713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,8192,0.02367288867632548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,12288,0.09500710831748115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,7168,0.02093155516518487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,10240,0.08082489172617595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,8192,0.06622222397062513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,6144,0.01832266648610433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,5120,0.016027554869651794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,7168,0.05788266658782959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,4096,0.013785777820481194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,6144,0.05066577924622429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,3584,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,5120,0.043086220820744835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,4096,0.03652088840802511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,3072,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,3584,0.033944000800450645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,3072,0.029370665550231934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,2048,0.008664000365469191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,2560,0.02605955633852217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,1536,0.007334222396214803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,2048,0.022708444131745234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,1024,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,1536,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,768,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,1024,0.016215110818545025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,512,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,768,0.01551466683546702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,256,0.0042239998777707415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,512,0.014130666851997375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,128,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,256,0.013374221821626028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,2560,0.010071111222108206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,128,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,32,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,64,0.012836444709036084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2560,32,0.012196444802814059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,65536,0.15390577581193712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,51200,0.1221768856048584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,65536,0.4687449137369792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2560,64,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,16384,0.04239822096294827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,51200,0.36901601155598956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,12288,0.032335999939176775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,16384,0.1148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,10240,0.026816000541051228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,12288,0.08775822321573894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,8192,0.02200444373819563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,10240,0.07458311319351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,7168,0.019720888800091214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,8192,0.060814221700032554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,6144,0.017633777525689866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,7168,0.053118222289615206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,5120,0.01519466605451372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,6144,0.04603200157483419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,4096,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,5120,0.040135112073686384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,3584,0.011884444289737277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,4096,0.033722665574815534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,3072,0.01074399964676963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,3584,0.03139111068513658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,2560,0.00959822204377916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,2048,0.007968000239796108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,3072,0.02706844442420536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,2560,0.024490666058328416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,2048,0.02146133283774058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,1536,0.018203554881943595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,1024,0.015797333584891427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,768,0.0052471111218134565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,768,0.01479288935661316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,512,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,512,0.013755555782053204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,256,0.004313777718279096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,1536,0.0070159998204973005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,128,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,256,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,64,0.003531555748648114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,1024,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,128,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,2048,32,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,64,0.01239111108912362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,65536,0.14101955625745985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,2048,32,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,51200,0.11074044307072957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,16384,0.03866844375928243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,51200,0.3633191055721707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,16384,0.11341688368055557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,12288,0.030099554194344416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,10240,0.025620443953408137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,12288,0.08637955453660752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,8192,0.02145333256986406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,10240,0.07315555546018812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,7168,0.01925244430700938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,8192,0.05934400028652615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,6144,0.016894222961531747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,7168,0.05199911197026571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,5120,0.0147724449634552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,6144,0.045294221904542714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,4096,0.012368888490729861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,5120,0.038753777742385864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,3584,0.010844444235165914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,65536,0.46225420633951825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,3072,0.009705777797434065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,3584,0.029721776644388836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,2560,0.008339555727110969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,3072,0.027238221632109746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,2048,0.007296889192528195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,2560,0.023660444551044043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,1536,0.006317333214812809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,2048,0.020959110723601446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,1024,0.0053173332578606075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,1024,0.015437333120240105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,768,0.005056888692908817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,4096,0.03275022241804335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,512,0.004531555705600315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,768,0.014510222607188754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,256,0.003944000022278892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,512,0.013426666458447775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,128,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,256,0.012797333300113678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,64,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,128,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1536,32,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,1536,0.017553778158293832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,64,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,65536,0.08409421973758274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1536,32,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,51200,0.06829244560665555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,16384,0.02730488942729102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,16384,0.1032515565554301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,51200,0.33168355623881024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,12288,0.02195377813445197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,10240,0.018927110566033255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,12288,0.07901422182718913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,8192,0.01607822212908003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,10240,0.06691644589106242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,7168,0.014781332678265043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,8192,0.05394311083687676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,6144,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,7168,0.047071112526787647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,6144,0.04157333241568671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,5120,0.011510222322411008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,4096,0.010617777705192566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,65536,0.4220595624711778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,5120,0.03567022085189819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,3584,0.009674666656388177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,3072,0.008328888979223039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,3584,0.028192000256644353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,2560,0.007332444190979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,4096,0.03015111221207513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,3072,0.02536800007025401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,2048,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,1536,0.0063075555695427795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,2560,0.02201777696609497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,1024,0.004876444323195351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,2048,0.019998222589492798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,768,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,1536,0.01719911065366533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,1024,0.014677332507239448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,512,0.0042568888101312844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,256,0.0038355555799272326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,768,0.01421955558988783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,128,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,256,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,64,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,128,0.012144889268610211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,1024,32,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,64,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,32,0.011530666715568967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,65536,0.058993776639302574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,51200,0.04799555407630073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,65536,0.4192035463121202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,1024,512,0.01295555548535453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,16384,0.020374221934212577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,12288,0.021167111065652635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,51200,0.33009423149956596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,16384,0.10241955518722534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,10240,0.01667555504375034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,8192,0.014114666316244336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,12288,0.0784319970342848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,7168,0.012776888906955719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,10240,0.06608177555931939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,6144,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,8192,0.052816000249650746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,5120,0.010072888599501716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,7168,0.04674399892489115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,6144,0.040920890039867826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,4096,0.010453333457310995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,3584,0.007416889071464539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,4096,0.029299557209014893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,5120,0.03467466764979892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,3072,0.0069431112044387394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,2560,0.0064293332397937775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,3584,0.027448889282014634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,2048,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,2560,0.021946666969193354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,3072,0.024703111913469102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,1536,0.004943111290534337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,1024,0.00425511101881663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,2048,0.01973511113060845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,768,0.004260444392760594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,1024,0.014161777165200977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,512,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,768,0.013815999858909182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,256,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,512,0.012823999755912356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,128,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,256,0.012200000385443369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,64,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,128,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,768,32,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,64,0.011490667031870948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,65536,0.04766311248143514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,32,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,51200,0.04021599888801575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,768,1536,0.016893333858913846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,16384,0.016328000360065036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,65536,0.38271199332343203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,51200,0.3019404411315918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,12288,0.016717332932684157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,10240,0.014974221587181091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,16384,0.09335644377602471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,12288,0.07079911231994629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,8192,0.014542222023010254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,10240,0.05980000231001112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,8192,0.04759022262361315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,6144,0.010067555639478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,7168,0.04204888807402717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,5120,0.009126221968068017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,6144,0.03740888833999634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,4096,0.008013333711359236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,5120,0.03232355581389533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,4096,0.027160889572567407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,3584,0.025436444414986506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,3072,0.006698666761318843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,7168,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,3072,0.022481777601771887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,2560,0.006511111226346757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,2048,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,2560,0.020672889219390023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,2048,0.018266666266653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,1536,0.0052355556852287715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,3584,0.007305777735180325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,1024,0.004290666845109728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,1536,0.015799999237060547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,768,0.0041759999261962045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,1024,0.014025777578353882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,512,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,768,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,256,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,512,0.012470222181744046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,256,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,128,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,64,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,128,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,512,32,0.0029333333174387612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,64,0.011482666763994427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,65536,0.03167466653717889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,512,32,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,51200,0.026160889201694067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,65536,0.38207377327813047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,16384,0.011619555453459421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,16384,0.09256355630026923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,12288,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,10240,0.010583110981517367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,12288,0.0704079998864068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,8192,0.009735999835862054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,10240,0.058205333020952016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,7168,0.009377777576446533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,8192,0.046979556481043495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,7168,0.04082399937841628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,6144,0.009036444127559662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,51200,0.301167991426256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,5120,0.00868088917599784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,4096,0.007668444679843054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,6144,0.03664888938268026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,5120,0.0313937763373057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,3584,0.0069297779765393995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,3072,0.0068693334857622785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,4096,0.027190221680535212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,2560,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,3584,0.02513244416978624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,3072,0.022421333524915908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,2048,0.005295111073387994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,1536,0.004576000074545543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,2560,0.02049422264099121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,1024,0.0041991112132867174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,2048,0.01830844415558709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,768,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,1536,0.015697777271270752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,512,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,1024,0.013818666338920593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,768,0.013080000049538083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,512,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,64,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,256,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,256,32,0.003030222323205736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,128,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,64,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,65536,0.024791111548741657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,256,32,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,51200,0.016227554943826463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,16384,0.011367999845080905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,65536,0.3816666603088379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,16384,0.09195733070373535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,51200,0.30068355136447483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,12288,0.010023111270533668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,10240,0.009336888790130615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,8192,0.009711999860074785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,12288,0.07015111049016316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,10240,0.058640890651279025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,7168,0.009340444372759925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,6144,0.00867555538813273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,7168,0.04041688972049289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,5120,0.008311111066076491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,6144,0.03632888860172696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,4096,0.007296889192528195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,5120,0.03136266602410211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,3584,0.0069546666410234236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,4096,0.026770666241645813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,3072,0.006268444574541516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,3584,0.025120000044504803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,2560,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,3072,0.02274133265018463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,2048,0.004908444566859139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,2560,0.020322667227851022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,1536,0.004552000098758274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,2048,0.018556444181336295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,1024,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,1536,0.01570755574438307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,768,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,1024,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,512,0.0032968889507982465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,768,0.012837332983811697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,256,0.0031386667655573953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,512,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,256,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,128,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,128,0.011676444775528379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,64,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,64,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,8192,0.046431110964881055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,128,32,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,65536,0.022601778308550518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,128,32,0.010763555765151978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,51200,0.014448000325096978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,65536,0.38139910168117946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,16384,0.009348444640636444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,51200,0.30093333456251353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,12288,0.009661333428488838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,16384,0.09172711107465957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,10240,0.008999110923873054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,12288,0.06975911060969035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,8192,0.00830577810605367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,10240,0.057870222462548145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,7168,0.007941332956155142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,8192,0.04569777846336365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,6144,0.009683555199040307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,7168,0.03999200132158067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,5120,0.008316444853941599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,6144,0.035608887672424316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,4096,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,5120,0.031303998496797346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,3584,0.006593777901596493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,4096,0.02676800058947669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,3072,0.022407111194398668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,3584,0.025495110286606684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,2560,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,2048,0.005132444616821078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,2560,0.020573332905769348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,2048,0.01824977828396691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,1536,0.004648000001907349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,1024,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,1536,0.01588000026014116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,768,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,1024,0.014136888914638095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,512,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,768,0.013094222380055321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,3072,0.006222222414281633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,256,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,128,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,512,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,256,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,128,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,64,32,0.002573333266708586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,64,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,64,32,0.010789333118332757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,51200,0.0199644449684355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,65536,0.38167911105685765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,51200,0.30071377754211426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,16384,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,12288,0.01687111126052009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,16384,0.09162844551934136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,10240,0.015116444892353482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,12288,0.06963466935687594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,8192,0.012805333567990197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,10240,0.05774133072959053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,65536,0.02384977704948849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,7168,0.01239911135700014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,8192,0.04600266615549723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,6144,0.014198222094111972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,5120,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,6144,0.0353724459807078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,7168,0.040397332774268255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,4096,0.00777422222826216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,5120,0.03125600020090739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,3584,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,3072,0.006611555400821898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,4096,0.02717599935001797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,3584,0.025103110406133864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,2560,0.005988444305128521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,3072,0.022732444935374792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,2048,0.005320888840489917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,1536,0.0046471113132105935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,2560,0.02030044462945726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,1024,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,2048,0.018266666266653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,768,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,1536,0.015855999456511605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,1024,0.013775111072593264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,512,0.003302222324742211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,768,0.013184000220563678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,128,0.0029288888391521242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,512,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,64,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,256,0.011578666667143503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,769,32,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,128,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,64,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,769,32,32,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,51200,2.4784569210476346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,51200,3.300657696194119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,16384,0.7990924517313639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,16384,1.0056550767686632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,12288,0.5963733461168077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,12288,0.7402497927347819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,16384,0.8168604638841418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,10240,0.5414302084181044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,12288,0.6214808887905544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,10240,0.6025564405653212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,51200,2.564242680867513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,8192,0.44994486702813047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,8192,0.48580535252888996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,10240,0.5249546898735894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,7168,0.38925332493252224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,8192,0.42664623260498047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,7168,0.426000886493259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,6144,0.3447191185421414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,7168,0.3780471218956842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,6144,0.36598843998379177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,5120,0.28656800587972003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,5120,0.30676976839701336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,6144,0.32966934310065377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,4096,0.22274398803710938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,4096,0.2499750985039605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,5120,0.28092177708943683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,3584,0.19924978415171304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,4096,0.23337690035502115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,3584,0.21929066711001924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,3072,0.16794310675726998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,3584,0.20675999588436553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,2560,0.14427467187245688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,3072,0.18447911739349365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,2560,0.16451288594139948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,2048,0.1182017723719279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,2560,0.15931467215220133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,2048,0.13645066155327693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,2048,0.1365893284479777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,3072,0.18944444921281603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,1536,0.09859644704394871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,1024,0.07024088833067152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,1536,0.10735911130905151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,1536,0.11177421940697564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,768,0.05625155236985949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,1024,0.08308800061543782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,1024,0.09000355667538112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,512,0.04075377848413255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,768,0.07668266693751018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,768,0.06849600209130181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,256,0.03126222226354811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,512,0.06604977448781331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,512,0.06009511152903239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,256,0.05504444572660658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,128,0.023052444060643513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,256,0.05577955643335978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,64,0.01979733341270023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,128,0.051757332351472646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,65536,32,0.020631111330456205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,64,0.05129422081841362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,65536,32,0.05129955543412102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,65536,128,0.04723644587728712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,65536,2.3027911716037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,51200,1.7950666215684679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,65536,3.2753378550211587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,16384,0.587548467848036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,51200,2.6325795915391708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,16384,0.7991058031717936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,12288,0.4311226738823785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,12288,0.5828462176852757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,16384,0.6565288967556423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,10240,0.36570578151279026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,12288,0.4991120232476129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,10240,0.5158070988125271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,65536,2.6094178093804254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,8192,0.292501343621148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,8192,0.39838843875461155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,51200,2.041243659125434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,7168,0.25040798717074925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,10240,0.42134843932257754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,6144,0.22172089417775473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,8192,0.34572267532348633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,7168,0.34208001030815977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,6144,0.30617332458496094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,7168,0.30600089497036403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,5120,0.19560621844397652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,4096,0.16186755233340794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,5120,0.2471004327138265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,6144,0.2680115434858534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,3584,0.13220977783203125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,4096,0.2018186648686727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,5120,0.22855999734666613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,4096,0.1903688907623291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,3072,0.113590227233039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,3584,0.17055200205908883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,3072,0.15352622667948404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,2560,0.10969866646660699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,3072,0.15137155850728354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,2048,0.07957866456773546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,2560,0.13310844368404812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,2560,0.1321964396370782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,2048,0.11123111512925889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,1536,0.0658444431093004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,2048,0.11329955524868435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,1024,0.047769778304629855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,1536,0.08770044644673665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,1536,0.0941306683752272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,1024,0.06846844487720065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,3584,0.1769697798622979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,768,0.03672800130314297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,1024,0.07632266812854342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,512,0.029135111305448744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,768,0.057003557682037354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,768,0.06578755378723145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,256,0.0207342223988639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,512,0.056840889983707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,512,0.05026400089263916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,128,0.019133332702848647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,256,0.04561422268549601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,256,0.0492524438434177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,64,0.015976000163290236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,51200,128,0.04252800014283922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,128,0.042945779032177396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,51200,32,0.01593155496650272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,64,0.04209422071774801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,51200,32,0.042246222496032715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,65536,0.8308382034301758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,65536,1.2574337853325737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,51200,0.6579546928405762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,51200,0.9993004269070096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,16384,0.22835111618041992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,65536,1.2671261893378365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,16384,0.30154577891031903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,12288,0.17478577295939127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,12288,0.22526489363776314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,51200,0.9969004525078667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,16384,0.32429244783189565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,10240,0.14496266841888428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,12288,0.2482551203833686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,10240,0.18988000022040474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,8192,0.12079555458492702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,10240,0.2114266686969333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,7168,0.10524444447623359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,8192,0.15435200267367893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,8192,0.17443554931216768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,7168,0.13736622863345677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,6144,0.09341600206163193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,7168,0.15617067284054226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,6144,0.1194346613354153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,5120,0.07800088988410102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,6144,0.13759644826253256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,4096,0.05972177452511258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,5120,0.10218489170074463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,5120,0.11904356214735244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,4096,0.08521422412660386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,3584,0.05166844526926676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,4096,0.10077155298656887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,3072,0.04491110973887973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,3584,0.0767093300819397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,3584,0.09154666794670953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,2560,0.038035555018319026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,3072,0.08220622274610731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,3072,0.06720444228914049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,2048,0.03278844555219015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,2560,0.07289066579606798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,2560,0.05893955628077189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,2048,0.0489671097861396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,1536,0.024996444582939148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,2048,0.06395644611782499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,1024,0.01832266648610433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,1536,0.03935111231274075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,1536,0.05474133292833964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,768,0.014752000570297241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,1024,0.03089155422316657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,1024,0.04565866788228353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,512,0.01256622208489312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,768,0.04079733292261759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,768,0.027859555350409612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,256,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,512,0.023024888502226934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,512,0.03666222095489502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,128,0.010007999837398529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,256,0.03113244308365716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,128,0.020263999700546265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,64,0.009480000370078618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,16384,128,0.02812888887193468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,16384,32,0.009717333647939894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,64,0.020026667250527274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,32,0.020054222808943856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,65536,0.5566417906019423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,16384,256,0.021298666795094807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,65536,1.0164160198635526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,51200,0.4590755568610297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,16384,0.15285510487026638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,16384,0.24732889069451225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,51200,0.812845336066352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,12288,0.10868177811304729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,12288,0.18584088484446207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,65536,1.0132524702284071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,10240,0.11068889167573716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,51200,0.7958301968044705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,12288,0.20072000556521943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,8192,0.08370577626758152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,10240,0.1565555599000719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,10240,0.1705155505074395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,8192,0.12770222293006048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,7168,0.07731022437413533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,8192,0.14152799712287054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,16384,0.2597777843475342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,6144,0.06738577948676215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,7168,0.1139635509914822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,6144,0.09870577520794338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,5120,0.05708977911207411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,6144,0.1130524476369222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,5120,0.0856613318125407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,4096,0.045160889625549316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,5120,0.09876355859968398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,4096,0.07162044445673625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,4096,0.08499111069573297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,7168,0.1269235610961914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,3584,0.0641075571378072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,3072,0.03494311041302151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,3584,0.0764897796842787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,3072,0.0688835581143697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,2560,0.030258665482203167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,3072,0.05452711052364773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,2560,0.0484808882077535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,2560,0.06227644284566244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,2048,0.02480711042881012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,3584,0.040207111173205905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,1536,0.019996444384257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,2048,0.05529510974884033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,2048,0.04184000028504265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,1024,0.015065777632925244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,1536,0.032925334241655134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,1536,0.04694844285647074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,768,0.012642666697502136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,1024,0.03972266780005561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,1024,0.027154665854242113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,512,0.009721777505344814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,768,0.023032888770103455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,768,0.036580443382263184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,256,0.007055111229419708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,512,0.03271289004219903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,512,0.020322667227851022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,128,0.006680000159475539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,256,0.018573333819707234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,256,0.029263112280103896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,64,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,12288,128,0.025778666138648987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,12288,32,0.006275555739800136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,128,0.018544000056054857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,64,0.017996443642510306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,12288,32,0.017502221796247695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,65536,0.5271849102444118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,65536,0.9078906377156576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,51200,0.42171645164489746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,51200,0.7223324245876737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,16384,0.13407999939388698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,65536,1.009584003024631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,16384,0.21919911437564424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,12288,0.10603466961118911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,51200,0.7967502276102701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,16384,0.25850576824612087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,12288,0.16570577356550428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,10240,0.09007733398013645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,10240,0.13947821987999812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,12288,0.19937777519226074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,8192,0.07326844665739271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,8192,0.11409778065151638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,8192,0.1409671174155341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,7168,0.06337689028845893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,7168,0.10140533579720391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,6144,0.05598044395446777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,7168,0.1266417768266466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,6144,0.08947110838360256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,5120,0.046358221107059054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,6144,0.11247732904222275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,10240,0.16981244087219238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,4096,0.03768800033463372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,5120,0.07576444413926867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,5120,0.09806844260957505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,4096,0.06396710872650146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,3584,0.0335804455810123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,4096,0.08345511224534775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,3072,0.030272887812720403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,3584,0.056976000467936196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,3584,0.07624000310897827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,2560,0.0251875552866194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,3072,0.04991822110282051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,3072,0.06877155436409844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,2048,0.02127733329931895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,2560,0.060953776041666664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,2560,0.04286666711171468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,2048,0.03693866729736328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,1536,0.017133333616786532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,2048,0.05456533365779453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,1024,0.01258666647805108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,1536,0.029861334297392104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,1024,0.0244835548930698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,1536,0.046624888976415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,768,0.010450666977299584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,1024,0.0388684438334571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,512,0.009510222408506606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,768,0.03611110978656345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,768,0.021420444051424663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,512,0.01863555610179901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,256,0.007324444750944774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,512,0.030847999784681533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,256,0.017281777328915067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,256,0.028174221515655518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,128,0.006304889089531369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,64,0.0074648890230390764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,128,0.016929777132140267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,10240,128,0.024899555577172175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,10240,32,0.007616889145639207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,32,0.016214221715927124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,65536,0.4203937848409017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,65536,0.7962222099304199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,51200,0.3222871091630724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,51200,0.629567994011773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,65536,0.8822355800204806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,16384,0.11408178011576335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,10240,64,0.016923555069499545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,16384,0.19323821862538657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,12288,0.08691377772225274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,12288,0.1469564437866211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,16384,0.22571110725402832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,10240,0.07283911440107557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,12288,0.17485154999627006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,10240,0.12495288583967422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,8192,0.05885510974460178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,10240,0.14958488941192627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,8192,0.102692445119222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,7168,0.05194222264819675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,8192,0.1242320007748074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,7168,0.09195199939939712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,6144,0.04408177733421326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,7168,0.1116480032602946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,6144,0.08052355713314481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,51200,0.694151136610243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,6144,0.09997066524293687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,5120,0.035762667655944824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,4096,0.030811554855770532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,5120,0.086681776576572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,5120,0.06739822361204359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,3584,0.027613333529896204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,4096,0.055480887492497764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,4096,0.07500800159242418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,3072,0.024719999896155462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,3584,0.05031288994683159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,3584,0.068031112353007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,2560,0.02111022174358368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,3072,0.04481155673662821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,3072,0.06213688850402832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,2048,0.018039999736679923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,2560,0.03905955619282193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,2560,0.05533155467775133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,1536,0.015012444721327888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,2048,0.033575110965304904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,1536,0.027656000521447923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,1024,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,1536,0.04298222064971924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,1024,0.02223644488387638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,1024,0.03642488850487603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,768,0.00907911111911138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,768,0.01925155520439148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,768,0.0321359998650021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,512,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,512,0.017287999391555786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,512,0.029919998513327703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,256,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,256,0.016142222616407607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,256,0.02764355474048191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,2048,0.05040622088644239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,128,0.006531555619504716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,64,0.006256000035338932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,8192,128,0.024919999970330134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,8192,32,0.006242666807439592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,64,0.015848888291252982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,32,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,65536,0.31526133749220103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,65536,0.738468435075548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,8192,128,0.01588355501492818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,51200,0.24860000610351562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,16384,0.09083466397391425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,51200,0.5883199903700087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,16384,0.17876356177859834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,65536,0.8819173177083334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,12288,0.07169422176149157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,16384,0.22590398788452148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,10240,0.06439910994635688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,12288,0.13513244522942439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,12288,0.1749253273010254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,8192,0.050662222835752696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,51200,0.6940613322787814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,10240,0.11481155289544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,10240,0.14895821942223444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,7168,0.046421332491768726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,8192,0.12380622492896186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,7168,0.0840337806277805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,6144,0.036997334824668035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,7168,0.1116195519765218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,5120,0.03421866562631395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,6144,0.07371289200252958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,6144,0.09984978040059407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,5120,0.061588446299235024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,4096,0.028183112541834515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,5120,0.08626400099860297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,4096,0.05097777644793192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,3584,0.025137777129809063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,4096,0.07455999983681573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,3584,0.04552533229192098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,3584,0.06779111093944974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,3072,0.022094221578703985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,8192,0.09479733308156331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,3072,0.04045866595374213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,3072,0.061462223529815674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,2560,0.03562666641341315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,2560,0.05549600058131748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,2048,0.015690666106012132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,2048,0.031480001078711614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,2048,0.04997155401441786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,1536,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,1536,0.02532177832391527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,1024,0.009855111440022787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,1024,0.020350222786267597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,2560,0.018630221486091614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,768,0.008420444197124904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,1024,0.034902221626705594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,768,0.03157511022355821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,768,0.018696889281272888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,512,0.0070257774657673305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,512,0.028852442900339764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,512,0.01696977847152286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,256,0.0063475556671619415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,1536,0.042359110381868154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,128,0.004947555561860402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,256,0.02632177703910404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,256,0.016006223029560514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,64,0.004640000147951974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,128,0.015236443943447538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,7168,128,0.02386933399571313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,7168,32,0.00462311092350218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,32,0.014502222339312235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,7168,64,0.015503111812803479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,65536,0.30862755245632595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,51200,0.22976443502638075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,65536,0.6829244295756022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,51200,0.5411573515997993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,16384,0.08451644579569499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,65536,0.7548951043023003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,16384,0.16566844781239828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,12288,0.06523555517196655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,51200,0.5960764355129665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,16384,0.19419644938574898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,12288,0.12644444571601018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,10240,0.05410222212473551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,10240,0.1072453326649136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,12288,0.15095644527011448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,8192,0.04253777861595154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,10240,0.12982489003075495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,7168,0.03911022345225016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,8192,0.08850133419036865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,8192,0.10873599847157796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,6144,0.03440444336997138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,7168,0.09779910908804999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,6144,0.06919644276301067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,5120,0.02943555514017741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,6144,0.08814577923880683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,5120,0.057670222388373486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,5120,0.07650399870342679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,4096,0.02407022151682112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,4096,0.06625955634646945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,4096,0.047594666481018066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,3584,0.021799999806616042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,3584,0.06053421894709269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,3072,0.019349333312776353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,7168,0.07969066831800672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,3072,0.038021332687801786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,2560,0.016381333271662395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,3072,0.05540000067816841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,2560,0.03398044572936164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,2048,0.014318222800890604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,2560,0.05054044392373827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,2048,0.02987733483314514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,2048,0.04589510957400004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,1536,0.011959111524952782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,3584,0.04317422376738655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,1536,0.039247112141715154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,1024,0.009329777624871995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,1536,0.024476443727811176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,768,0.008008889026112026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,1024,0.019920888874265883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,1024,0.03199288911289639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,512,0.006590222318967183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,768,0.029895110262764826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,512,0.0158915552828047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,768,0.01789688898457421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,256,0.005449777675999536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,512,0.027504000398847792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,128,0.004918222212129169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,256,0.025184889634450276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,256,0.015531554818153381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,64,0.004562666846646203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,128,0.014843554960356818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,6144,128,0.02281155520015293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,6144,32,0.004505777938498391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,64,0.014949333336618213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,6144,32,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,65536,0.2837679915957981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,65536,0.6280657980177138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,51200,0.2143688864178128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,51200,0.492578665415446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,16384,0.069504890176985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,65536,0.7535546620686849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,16384,0.15381156073676214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,51200,0.5930275387234157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,12288,0.053598221805360585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,16384,0.19393244054582384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,10240,0.04807110958629184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,12288,0.1178471114900377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,12288,0.15083732869890001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,8192,0.037760890192455716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,10240,0.10061066680484348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,10240,0.12913155555725098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,8192,0.08288533157772489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,8192,0.10910577906502618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,7168,0.0739591121673584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,6144,0.029841777351167467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,7168,0.09796888960732354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,6144,0.0639795528517829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,5120,0.02536800007025401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,6144,0.08796622355779012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,5120,0.05346399876806471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,5120,0.07672888702816434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,4096,0.021335111724005804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,7168,0.034930666287740074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,3584,0.02012355625629425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,4096,0.06614933411280315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,4096,0.04563022322124905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,3072,0.018194667167133756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,3584,0.06063822242948744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,3072,0.03642222285270691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,3072,0.05529777871237861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,2560,0.015981333123313058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,2560,0.031856000423431396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,2560,0.05023466547330221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,2048,0.01347733371787601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,2048,0.045187556081347995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,2048,0.028357333607143823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,1536,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,3584,0.040957334968778826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,1536,0.022986667023764715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,1536,0.037751999166276716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,1024,0.03153866529464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,1024,0.01891644464598762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,768,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,768,0.029137776957617864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,768,0.01688888834582435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,512,0.006896889044178857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,512,0.02753866712252299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,512,0.015568888849682279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,256,0.0053324442770746015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,256,0.025073778298166063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,128,0.00517511119445165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,256,0.014892444014549255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,1024,0.00902844468752543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,64,0.004585777719815572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,128,0.014096000128322177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,5120,128,0.023037332627508376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,5120,32,0.006274666637182236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,64,0.014142221874660917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,5120,32,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,65536,0.19088621934254965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,51200,0.1500800053278605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,65536,0.5863199763827854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,51200,0.45804887347751194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,16384,0.0520924429098765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,65536,0.7531146473354764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,16384,0.14279022481706408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,12288,0.04028266668319702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,51200,0.5924844212002224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,16384,0.19368444548712838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,12288,0.10963022046618992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,10240,0.03753066725201077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,12288,0.1497288942337036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,8192,0.030581331915325586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,10240,0.0934071143468221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,10240,0.13005244731903076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,7168,0.027375110321574744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,8192,0.07696621947818331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,8192,0.10819555653466119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,6144,0.022848000129063923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,7168,0.09675911400053237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,6144,0.05890044238832262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,5120,0.01995466649532318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,6144,0.08717155456542969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,5120,0.049088888698154025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,5120,0.07558311356438531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,4096,0.016951110627916124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,4096,0.042140444119771324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,4096,0.06615822182761298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,3584,0.015704888436529372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,3584,0.03769422239727444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,3072,0.014408000641398959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,7168,0.06913155317306519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,3072,0.055461333857642285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,3072,0.034027556578318276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,2560,0.013040000365840064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,2560,0.02962577674123976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,2048,0.011483555866612328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,2560,0.05004177822007073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,2048,0.025642666551801894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,1536,0.009416888985368941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,2048,0.04529866576194763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,3584,0.06038933330112033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,1024,0.007330666813585494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,1536,0.0211173329088423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,1536,0.03708711266517639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,768,0.006616888774765863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,1024,0.03156977891921997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,1024,0.01790844400723775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,512,0.0052871112194326185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,768,0.02922399838765462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,768,0.016240888171725802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,256,0.005015111217896144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,512,0.027487110760476854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,256,0.025799999634424847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,256,0.014215999179416232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,128,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,4096,128,0.022658665974934895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,128,0.013793778088357715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,64,0.004315555509593752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,4096,32,0.004543111142185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,64,0.013801777528391944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,32,0.012930666406949362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,4096,512,0.014846222268210517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,65536,0.19530844688415527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,51200,0.14321245087517634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,65536,0.5463226636250814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,51200,0.4329555564456516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,16384,0.04820800158712599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,65536,0.7520826657613119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,16384,0.13464532958136663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,51200,0.5930977927313911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,12288,0.03703822361098395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,16384,0.1930462254418267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,10240,0.03238222334120009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,12288,0.1027066641383701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,12288,0.1503022246890598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,10240,0.08757955498165554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,8192,0.02646044393380483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,10240,0.1290373272365994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,7168,0.02369600037733714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,8192,0.0724124444855584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,8192,0.10885333352618748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,7168,0.06401866674423218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,6144,0.020813332663642038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,7168,0.0967164436976115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,5120,0.018315555320845712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,6144,0.05502044492297702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,6144,0.08746933274798924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,5120,0.046888887882232666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,4096,0.015591111448076038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,5120,0.07566844754748873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,3584,0.014173333843549093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,4096,0.039480000734329224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,4096,0.06583288643095228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,3584,0.03639199998643663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,3072,0.01276622215906779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,3584,0.05997777647442288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,2560,0.011159111228254108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,3072,0.03206844462288751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,3072,0.05519466598828634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,2048,0.009711999860074785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,2560,0.02886311213175456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,2560,0.049272000789642334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,1536,0.007967111137178209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,2048,0.025715554753939312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,2048,0.044328888257344566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,1024,0.006606222026877933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,1536,0.020284444093704224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,1536,0.03634577658441331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,1024,0.01717511150572035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,1024,0.031089779403474595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,768,0.029288887977600098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,768,0.015831111205948722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,512,0.0052560000783867305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,512,0.014467555615637036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,256,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,256,0.013796444568369122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,256,0.025841777523358662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,128,0.004252444538805219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,768,0.005892444401979446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,128,0.013449778159459433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,64,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,128,0.023285332653257582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3584,32,0.003931555483076307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3584,512,0.027126222848892212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,32,0.013239111337396832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,65536,0.17100177870856392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,65536,0.5140053431193033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,51200,0.1339306698905097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,65536,0.6249430974324545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3584,64,0.01332444449265798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,51200,0.40591200192769367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,16384,0.045336001449161105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,51200,0.49294577704535586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,16384,0.12514844205644396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,12288,0.03432622220781114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,16384,0.16062133842044407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,10240,0.030685332086351182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,12288,0.09637955824534099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,12288,0.12531999746958414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,8192,0.02479644450876448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,10240,0.08155822091632418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,8192,0.06732355886035495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,10240,0.10905866490470038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,7168,0.02184088859293196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,8192,0.0913262234793769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,6144,0.01944977707333035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,7168,0.08172800143559773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,7168,0.05951288673612806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,5120,0.016576000385814242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,6144,0.05076977941724989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,6144,0.07395644320382012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,4096,0.01422044469250573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,5120,0.06579999791251288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,5120,0.043453332450654775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,3584,0.012869333227475485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,4096,0.03712177938885159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,4096,0.0575795571009318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,3072,0.011844444606039258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,3584,0.05225244495603773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,3072,0.030215998490651447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,3584,0.03429422113630507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,2560,0.010508444574144153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,3072,0.04734844300482008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,2048,0.008647999829716152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,2560,0.043581333425309926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,2560,0.026595556073718604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,2048,0.023690667417314317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,1536,0.007631111476156447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,2048,0.038062221474117704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,1024,0.005952888892756567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,1536,0.019262222780121695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,1536,0.03299199872546726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,1024,0.016559999850061204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,1024,0.02889333168665568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,768,0.005608888963858287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,512,0.004621333132187526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,768,0.014839111102951897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,768,0.027033777700530157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,256,0.0042071110672420925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,512,0.02540622154871623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,512,0.014283556077215405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,256,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,128,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,256,0.023785778217845496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,64,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,128,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,3072,128,0.022699554761250813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,3072,32,0.003607999947335985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,64,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,65536,0.15251378218332926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,3072,32,0.013012444807423485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,51200,0.11274666256374782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,65536,0.5018044577704536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,65536,0.6215333408779568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,51200,0.39597511291503906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,51200,0.49195199542575413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,16384,0.12371644708845352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,12288,0.03039466672473484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,16384,0.16080355644226074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,12288,0.09488622347513835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,12288,0.12575111124250624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,10240,0.025643555654419795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,10240,0.08026755518383451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,16384,0.041747556792365186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,10240,0.10866844654083252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,8192,0.06602311134338379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,8192,0.09111733569039239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,7168,0.01924622224436866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,7168,0.08172711398866442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,7168,0.05853155586454603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,6144,0.017073776986863878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,6144,0.07346310880449083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,6144,0.050108446015252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,5120,0.015150222513410779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,8192,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,5120,0.06500799788369073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,5120,0.042988446023729115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,4096,0.012936000194814471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,3584,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,4096,0.03647644321123759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,4096,0.05721600188149346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,3584,0.03336266676584879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,3072,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,3584,0.05003289050526089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,2560,0.009434666898515489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,3072,0.028935998678207397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,3072,0.04673688941531711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,2048,0.008175111479229396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,2560,0.026022222306993272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,2560,0.04189777705404493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,2048,0.037419554260041975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,2048,0.0236799998415841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,1536,0.007017778025733099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,1536,0.03242044316397773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,1024,0.006203555398517185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,1024,0.016160888804329764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,1024,0.028359111812379625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,768,0.005631111148330901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,768,0.014831110835075378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,768,0.026920888159010146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,512,0.004944889081848992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,1536,0.01921244462331136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,512,0.013815999858909182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,256,0.0042239998777707415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,512,0.02581600017017788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,128,0.003944888710975647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,256,0.024043555061022442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,256,0.013282666603724161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,64,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2560,128,0.021426666114065383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,128,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2560,32,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,64,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,65536,0.13436267111036512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2560,32,0.012810666528013019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,51200,0.10348800155851577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,65536,0.4656702147589789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,51200,0.3676906691657172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,65536,0.6203911039564345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,16384,0.036228444841172956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,51200,0.4904426468743219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,16384,0.1606488890118069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,12288,0.0277493331167433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,12288,0.12513689200083414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,12288,0.08794489171769883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,10240,0.023014222582181294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,10240,0.1074577768643697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,10240,0.07443910837173462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,8192,0.019288889235920377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,8192,0.06105955441792806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,8192,0.09008889065848456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,7168,0.017089777522616915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,16384,0.1142231093512641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,6144,0.015203555425008139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,7168,0.05319111214743721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,7168,0.08173866404427423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,6144,0.04587911234961616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,5120,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,6144,0.07326489024692111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,4096,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,5120,0.039784000979529485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,5120,0.0650906695259942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,3584,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,4096,0.03382133444150289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,4096,0.056553781032562256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,3072,0.009687999884287516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,3584,0.05190755592452156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,3584,0.03122577733463711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,2560,0.008678221868144141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,3072,0.02751733362674713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,3072,0.04764088988304138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,2048,0.007307555940416124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,2560,0.041161778900358416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,2560,0.02424000038041009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,1536,0.00656088896923595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,2048,0.021407110823525324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,2048,0.03741066654523214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,1024,0.005631111148330901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,1536,0.01824977828396691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,1536,0.032291554742389254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,768,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,1024,0.015558222929636637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,1024,0.028768890433841284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,768,0.014198222094111972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,512,0.004582222137186262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,512,0.013448889056841532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,512,0.025781333446502686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,256,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,256,0.0243857784403695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,256,0.013176889055305056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,128,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,128,0.02158755560715993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,128,0.012535110943847232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,64,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,2048,32,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,64,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,2048,32,0.012130666938092975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,2048,768,0.02680711117055681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,65536,0.11295910676320393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,51200,0.08901955684026082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,65536,0.42705244488186306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,51200,0.3374337885114882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,16384,0.03145066565937466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,65536,0.6097884707980686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,16384,0.1575253274705675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,51200,0.4819839795430501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,16384,0.10508533318837483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,12288,0.023755555351575215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,10240,0.020655110478401184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,12288,0.07981777853435941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,12288,0.12291555934482151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,10240,0.06798222329881456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,8192,0.017268444101015728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,10240,0.1063422229554918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,8192,0.055023110575146146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,7168,0.015252444479200574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,8192,0.08935022354125977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,7168,0.04903644323348999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,6144,0.013591110706329346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,7168,0.08037955231136747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,6144,0.042252444558673434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,5120,0.011980444192886353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,6144,0.0725039972199334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,5120,0.03651288814014859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,4096,0.010407999985747868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,5120,0.0637306637234158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,3584,0.009388444324334463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,4096,0.030786666605207656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,4096,0.055904891755845815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,3584,0.02888444397184584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,3072,0.008671999805503422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,3584,0.05102133419778612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,2560,0.007693332930405934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,3072,0.046859555774264865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,2560,0.02270222206910451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,2048,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,2560,0.040475553936428495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,2048,0.019928889142142404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,2048,0.03672711054484049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,1536,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,1536,0.03198755449718899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,1536,0.016908443636364408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,1024,0.005128889034191768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,1024,0.014846222268210517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,1024,0.028252445989184912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,768,0.004615111069546806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,768,0.02650044361750285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,768,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,512,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,3072,0.025756445195939805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,256,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,512,0.012777778009573618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,512,0.02514755560292138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,128,0.003547555456558863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,256,0.024030221833123103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,256,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,64,0.003168888803985384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,128,0.012159999873903064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1536,32,0.0032559999575217566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,32,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1536,64,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,65536,0.0756817791197035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,65536,0.4210231039259169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,51200,0.06087555487950643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,65536,0.6102524333530003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1536,128,0.020977778567208186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,16384,0.025955556167496577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,51200,0.3314880000220405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,16384,0.10397511058383518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,12288,0.02296177711751726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,51200,0.4816177686055501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,16384,0.15762666861216226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,12288,0.07856088876724243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,10240,0.018262222409248352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,12288,0.12229243914286296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,8192,0.016202666693263583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,10240,0.06679289208518134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,10240,0.10539733039008246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,8192,0.05347555544641283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,7168,0.013776000175211163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,8192,0.08833777904510498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,6144,0.01311466677321328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,7168,0.047247999244266085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,7168,0.08014044496748182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,6144,0.04108622339036729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,5120,0.010975110861990186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,5120,0.035283555587132774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,4096,0.008011555506123437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,5120,0.06306666798061795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,4096,0.029567999972237483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,4096,0.054917332198884755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,3584,0.007654222349325816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,3584,0.04827644427617391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,3072,0.0069288888739215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,3584,0.02795555525355869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,3072,0.02444444431198968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,3072,0.04531911015510559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,2560,0.006678222368160884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,6144,0.07155911127726237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,2048,0.005889777921968036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,2560,0.04008177916208903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,2560,0.022216889593336318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,1536,0.0052737775776121355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,2048,0.019667555888493855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,2048,0.03641422258483039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,1024,0.004599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,1536,0.03161866797341241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,1536,0.016937777400016785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,768,0.004211555752489301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,1024,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,1024,0.027855998939938013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,768,0.02645688917901781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,768,0.013578666581047906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,512,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,512,0.024753777517212763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,256,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,512,0.012835555606418185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,256,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,256,0.02374222212367588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,128,0.003539555602603488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,1024,128,0.02108000053299798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,64,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,128,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,1024,32,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,64,0.01180355581972334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,65536,0.05867111020618015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,1024,32,0.012055111428101858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,65536,0.3827048937479655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,51200,0.04785155587726169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,65536,0.6104764408535427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,16384,0.01925155520439148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,51200,0.30192356639438206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,51200,0.4816880226135254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,16384,0.09453688727484809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,16384,0.15730667114257812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,12288,0.017960889471901786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,10240,0.015311110350820752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,12288,0.12196711699167888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,12288,0.07208000289069282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,8192,0.013250666360060373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,10240,0.06102489100562202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,10240,0.1054604450861613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,7168,0.012040889097584618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,8192,0.048320889472961426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,8192,0.08798578050401475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,6144,0.010649777948856354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,7168,0.042837331692377724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,7168,0.07981155316034953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,5120,0.00942311104800966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,6144,0.03747111227777269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,5120,0.03333244389957852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,5120,0.06269244352976482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,4096,0.008040000167157915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,4096,0.02776622275511424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,4096,0.05428533421622383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,3584,0.02584533393383026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,6144,0.07131822241677178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,3584,0.049040890402264066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,3072,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,3072,0.04396177662743462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,3072,0.023015111684799194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,2560,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,2560,0.020608888732062448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,2048,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,2048,0.018917333748605516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,2048,0.036658667855792575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,1536,0.005232000102599462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,1536,0.0162426663769616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,3584,0.006986666884687211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,1024,0.0042568888101312844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,1536,0.03156800071398417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,1024,0.014132445057233175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,1024,0.027882665395736694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,768,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,768,0.013134222063753339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,768,0.025811556312772963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,512,0.003608889049953885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,512,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,256,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,512,0.025409777959187824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,256,0.012138666378127204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,256,0.023178666830062866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,128,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,128,0.02109333376089732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,128,0.011870221959220039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,64,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,768,32,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,64,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,768,32,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,65536,0.04394577940305074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,65536,0.38184356689453125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,51200,0.03732088870472378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,768,2560,0.04041599896219041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,65536,0.6097564697265625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,16384,0.017062221964200337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,51200,0.30158400535583496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,51200,0.4812409083048503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,16384,0.09299911393059625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,12288,0.015199111567603217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,16384,0.15677600436740452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,10240,0.01482222146458096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,12288,0.07079733080334134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,10240,0.05927466683917575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,8192,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,10240,0.10523288779788548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,8192,0.047747555706236094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,7168,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,8192,0.0882951087421841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,7168,0.04238311118549771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,6144,0.010250666903124915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,7168,0.08018044630686443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,6144,0.03689510954750909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,5120,0.00868711123863856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,6144,0.07147377729415894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,12288,0.12229511472913955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,4096,0.007692444655630324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,5120,0.03161244591077169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,5120,0.062317331631978355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,4096,0.02719288898838891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,3584,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,4096,0.05273688832918803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,3072,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,3584,0.025507556067572698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,3584,0.04834489027659098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,3072,0.022660444180170696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,2560,0.005863111052248213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,3072,0.043947554296917386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,2048,0.005320888840489917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,2560,0.020439111524158053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,2560,0.03976444403330485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,1536,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,2048,0.018575999471876357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,2048,0.03606399893760681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,1024,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,1536,0.03126044405831231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,1536,0.016081778539551627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,1024,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,768,0.004200889004601372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,1024,0.027480888697836135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,512,0.0038453332251972626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,768,0.012843555874294706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,768,0.025850666893853083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,256,0.003370666669474708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,512,0.012158222496509552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,512,0.024736889534526404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,128,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,256,0.023003555006451074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,256,0.011795555551846823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,128,0.011490667031870948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,64,0.003010666618744532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,512,128,0.02033333314789666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,512,32,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,65536,0.03333688775698344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,64,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,512,32,0.011121778024567498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,51200,0.02732177740997738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,65536,0.3816800117492676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,51200,0.30099466111924916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,16384,0.012724444270133972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,65536,0.609115547604031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,16384,0.0924346645673116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,16384,0.15678932931688097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,12288,0.01072177787621816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,51200,0.48158401913113064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,10240,0.010016000105275048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,12288,0.07022222545411852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,12288,0.12207199467553033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,8192,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,10240,0.058628446525997586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,10240,0.10582488775253296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,7168,0.008307555483447181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,8192,0.04713066750102573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,8192,0.08763555685679118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,7168,0.04182577795452542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,7168,0.07833955685297649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,6144,0.009735110733244155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,5120,0.008656000097592672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,6144,0.035996443695492215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,6144,0.06880888673994276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,4096,0.007333333293596904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,5120,0.0321688883834415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,5120,0.060306668281555176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,3584,0.006990222467316522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,4096,0.027114666170544092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,4096,0.05245866709285312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,3072,0.006680000159475539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,3584,0.02511288887924618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,3584,0.04769600099987454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,2560,0.0059262220230367445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,3072,0.044017778502570264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,3072,0.022688888841205176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,2560,0.020644444558355544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,2048,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,2560,0.039805332819620766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,1536,0.004858666823969947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,2048,0.018252443936136033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,2048,0.03640710976388719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,1024,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,1536,0.03122933374510871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,1536,0.015821332732836407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,768,0.003921777837806278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,1024,0.013879111243618859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,768,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,512,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,768,0.02606222199069129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,512,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,256,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,512,0.024715556038750544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,256,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,256,0.023025777604844835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,128,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,128,0.02071199980047014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,64,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,64,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,256,32,0.0030666666312350165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,256,32,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,65536,0.02586311101913452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,256,1024,0.027433777848879497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,65536,0.38048532274034286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,51200,0.01744622157679664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,51200,0.3002844386630588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,65536,0.6096248626708984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,16384,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,16384,0.09230755435095893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,51200,0.48026757770114475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,12288,0.010082667072614035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,16384,0.15798666742112902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,12288,0.06972177823384602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,10240,0.00904622260067198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,12288,0.12188177638583714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,10240,0.05789955457051595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,8192,0.008346666892369589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,10240,0.10505600108040704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,7168,0.008025777836640676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,8192,0.04680888851483663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,7168,0.04071199893951416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,6144,0.007657777931955125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,7168,0.07786577939987183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,6144,0.03574577636188931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,5120,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,5120,0.03150577677620782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,5120,0.06000444624159071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,4096,0.007275555696752336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,8192,0.08773422241210938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,4096,0.026898667216300964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,4096,0.052433778842290245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,3584,0.006693333387374878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,3584,0.025117332736651104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,6144,0.06842488712734647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,3072,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,3072,0.022320000661744013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,3072,0.043679998980628125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,2560,0.005574222240183089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,2560,0.02032977839310964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,2560,0.03942044576009115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,2048,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,2048,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,2048,0.03609955641958449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,1536,0.004579555657174853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,1536,0.015853333804342482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,1536,0.03128711051411099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,1024,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,3584,0.04768088791105482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,768,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,1024,0.028655999236636694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,768,0.012900444368521372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,512,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,768,0.026920000712076824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,512,0.012246222131782107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,512,0.024679111109839544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,256,0.003113777687152227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,256,0.011935110721323224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,256,0.023136888941129048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,128,0.011924444801277585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,768,128,128,0.020793777373101976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,64,0.0027875554644399216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,128,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,1024,0.013787555197874704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,64,0.011544889046086205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,65536,0.02474755545457204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,128,32,0.011165333290894827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,65536,0.3809102111392551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,51200,0.30030043919881183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,16384,0.00945688866906696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,12288,0.008025777836640676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,16384,0.09225600295596653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,12288,0.06936799817615084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,10240,0.008712000317043727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,8192,0.00795911086930169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,10240,0.0576711098353068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,51200,0.01610577768749661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,7168,0.008019555773999956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,6144,0.007343110938866933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,7168,0.04028444488843282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,6144,0.035227557023366295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,5120,0.007801777786678738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,4096,0.00739733295308219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,5120,0.031201776531007554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,3584,0.00702044450574451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,3584,0.02498488956027561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,3072,0.006657777975002925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,3072,0.022508444057570562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,2560,0.006605333338181178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,2560,0.02033422225051456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,2048,0.005032888717121548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,2048,0.01834133267402649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,1536,0.004587555511130227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,1536,0.015871110889646742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,8192,0.04691288868586222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,1024,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,1024,0.01367911116944419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,768,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,768,0.01275644451379776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,4096,0.02679288884003957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,512,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,256,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,512,0.01182577759027481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,128,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,256,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,64,0.0025635556214385563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,128,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,64,32,0.002558222247494592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,64,0.011038222246699862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,65536,0.02641244398223029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,64,32,0.010770666930410596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,51200,0.021436444587177698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,65536,0.3809902138180203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,16384,0.010384000009960597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,51200,0.3002684381273058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,12288,0.00904266701804267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,16384,0.09204977750778198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,10240,0.008023999631404877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,12288,0.06970133384068807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,8192,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,10240,0.0581279993057251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,7168,0.007697777615653143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,8192,0.044991112417644925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,6144,0.007305777735180325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,7168,0.04149511125352647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,5120,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,6144,0.03530222177505493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,4096,0.007344000041484833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,5120,0.031146665414174397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,3584,0.00701333334048589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,4096,0.02681244413057963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,3072,0.006698666761318843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,3584,0.02511466708448198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,2560,0.006599111275540457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,3072,0.022322666313913133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,2048,0.004970666848950916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,2560,0.020336000455750358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,1536,0.004587555511130227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,2048,0.018205333087179396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,1024,0.0038311113086011675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,1536,0.01556622154182858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,768,0.003531555748648114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,1024,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,512,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,768,0.012626666989591388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,256,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,512,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,256,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,64,0.002577777744995223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,128,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,768,32,32,0.0025635556214385563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,64,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,768,32,32,0.010791111323568555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,51200,1.858618630303277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,16384,0.610023127661811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,51200,2.469317330254449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,12288,0.4633013407389323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,16384,0.7664471202426486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,12288,0.5502008861965603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,10240,0.3861680030822754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,10240,0.4510444535149469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,8192,0.314733346303304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,7168,0.273580445183648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,8192,0.3623662259843614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,6144,0.23704089058770073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,7168,0.317068444357978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,5120,0.20068800449371338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,6144,0.2779573334587945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,5120,0.23474756876627603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,4096,0.18244355254703096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,3584,0.16153777970208064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,3072,0.13992888397640652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,4096,0.19122666782803008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,3584,0.16634933153788248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,2560,0.11919200420379639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,2560,0.126111110051473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,2048,0.09757688972685073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,1536,0.07612444294823541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,1024,0.05482933256361219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,2048,0.10401333040661281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,1536,0.0833031137784322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,768,0.044581333796183266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,512,0.033971554703182645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,1024,0.06458844741185506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,256,0.02160533269246419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,768,0.05412266651789347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,128,0.01958755486541324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,512,0.04760533240106371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,256,0.04222755630811056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,64,0.016923555069499545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,65536,32,0.018912000788582694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,128,0.03997777899106344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,64,0.03869955407248603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,3072,0.1454577843348185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,65536,32,0.03847377830081516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,51200,1.2847750981648762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,65536,1.638511127895779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,16384,0.4477626482645671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,65536,2.4757323794894748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,12288,0.3171093463897705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,16384,0.5882942411634657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,10240,0.3111084567175971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,51200,1.9664221869574652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,12288,0.43816089630126953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,8192,0.2482551203833686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,10240,0.3597377671135797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,7168,0.2049920029110379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,8192,0.2947022120157878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,6144,0.17943199475606283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,7168,0.2564355532328288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,6144,0.22270578808254668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,5120,0.15135378307766384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,4096,0.13110310501522487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,3584,0.10813422335518731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,5120,0.18870577547285292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,3072,0.10002844201193915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,3584,0.13528445031907824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,2560,0.08276444673538208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,3072,0.11826933754814996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,2048,0.06749066379335192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,2560,0.10284444358613755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,1536,0.05053955647680494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,2048,0.08543022473653157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,1024,0.03682933251063029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,1536,0.06864444414774577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,4096,0.15353688928816053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,768,0.03039466672473484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,512,0.022737777895397607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,1024,0.05412711037529839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,768,0.0452133317788442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,256,0.016501333978441026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,128,0.014074666632546319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,256,0.03495644529660543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,64,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,128,0.033398220936457314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,64,0.03228444523281521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,51200,32,0.011848889291286469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,32,0.03226044442918565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,65536,0.5830995771620009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,51200,512,0.03921511107020908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,51200,0.4530960188971625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,65536,0.9179279539320203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,16384,0.1559013393190172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,51200,0.7340097957187228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,12288,0.11443732844458686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,16384,0.2240604559580485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,12288,0.17073688242170545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,10240,0.09694844484329224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,10240,0.1434239943822225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,8192,0.07971644401550293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,7168,0.07007110781139798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,6144,0.061963558197021484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,8192,0.11719644069671631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,5120,0.05535644292831421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,6144,0.09173422389560276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,5120,0.07946044206619263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,4096,0.06480355395211114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,3584,0.040256887674331665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,3584,0.059416890144348145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,3072,0.035467555125554405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,3072,0.05302400059170193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,4096,0.0443173348903656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,2560,0.03080000148879157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,2048,0.025829333398077223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,2560,0.04592000113593208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,1536,0.02091111077202691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,2048,0.0380711125002967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,1024,0.016507556041081745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,768,0.014224888549910652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,1536,0.031146665414174397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,512,0.011366222467687396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,1024,0.025775111383861963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,256,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,768,0.022727999422285292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,7168,0.10447555780410767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,128,0.0069724445541699724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,64,0.006247111078765657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,128,0.017913776967260573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,16384,32,0.006605333338181178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,64,0.017874666386180453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,32,0.017605332864655387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,65536,0.504234684838189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,512,0.02019466625319587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,51200,0.39003554979960126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,65536,0.7481990920172797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,16384,256,0.018559111489189994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,16384,0.13336622714996338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,12288,0.09993333286709255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,16384,0.18446755409240723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,10240,0.08533066511154175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,12288,0.14077511098649767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,51200,0.591458691491021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,8192,0.0697751111454434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,7168,0.06106399827533298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,10240,0.1193973355823093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,8192,0.0978248914082845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,6144,0.05449155635303921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,5120,0.046498666206995644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,6144,0.07580000162124634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,4096,0.03825866513782077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,5120,0.06673866510391235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,3584,0.03401422169473436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,4096,0.0532915559079912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,3072,0.02994489007525974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,3584,0.048026667700873486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,2560,0.025210666987631056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,3072,0.042706665065553456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,2048,0.020983111527231004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,2560,0.039048890272776283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,1536,0.016398222910033334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,2048,0.03197866678237915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,1024,0.012936000194814471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,1536,0.027116444375779893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,7168,0.08615022235446507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,768,0.010692444112565784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,512,0.00832799987660514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,1024,0.022690667046440974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,256,0.005949333310127258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,768,0.020056888461112976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,128,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,512,0.018045332696702745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,64,0.004927111168702443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,256,0.016672000288963318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,12288,32,0.005218666460778978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,128,0.01649599936273363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,64,0.016186666157510545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,12288,32,0.015796444482273526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,65536,0.48497332466973203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,51200,0.3831760088602702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,16384,0.12682667043473986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,65536,0.6599662039015028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,12288,0.09723377890057033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,51200,0.5209529134962294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,12288,0.12547022766537136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,10240,0.0820488863521152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,8192,0.06729244523578219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,10240,0.10682755708694458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,8192,0.08754311005274455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,7168,0.05965955389870537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,6144,0.05156533254517449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,7168,0.0770711104075114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,5120,0.04458399944835239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,5120,0.05810578001870049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,4096,0.03644088904062907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,4096,0.04761066701677111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,3584,0.0322675539387597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,3584,0.04282755653063456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,3072,0.02822222312291463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,3072,0.03875911235809326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,6144,0.06806221935484144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,2560,0.02400622268517812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,2048,0.019689778486887615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,2560,0.03436533278889126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,1536,0.0162471118900511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,16384,0.16505511601765951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,2048,0.029027554723951552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,1024,0.013822221921549903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,768,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,1536,0.025040000677108765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,1024,0.020868443780475195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,512,0.008667555948098501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,256,0.006678222368160884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,768,0.01868266695075565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,128,0.005564444594913059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,512,0.016840888394249808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,64,0.004934222333961063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,256,0.01609599921438429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,128,0.015491555134455362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,10240,32,0.005221333354711533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,64,0.015480000111791821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,10240,32,0.014814222852389017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,65536,0.311772452460395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,51200,0.2664853201972114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,16384,0.0840524435043335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,65536,0.5757030910915798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,12288,0.06038844585418701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,51200,0.4582542313469781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,16384,0.14356533686319986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,10240,0.05067200130886502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,8192,0.04166311025619507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,12288,0.10999377568562825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,10240,0.09367822276221381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,7168,0.03768710957633124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,6144,0.032413333654403687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,8192,0.07736711369620429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,5120,0.028500444359249536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,7168,0.06810577710469563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,6144,0.05982222159703573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,5120,0.051073776351081006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,3584,0.021739555729760066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,4096,0.04211644331614176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,3072,0.018983999888102215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,3584,0.0381368895371755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,3072,0.03485066692034403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,2560,0.016884444488419425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,2560,0.03139822350607978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,2048,0.014676445060306124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,2048,0.026100445124838088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,4096,0.023696889479955036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,1536,0.012479110724396177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,1024,0.010126222338941362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,1536,0.022859555151727464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,768,0.008283555507659912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,1024,0.018725333942307364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,512,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,768,0.017239999439981248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,256,0.00556355549229516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,512,0.01612444387541877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,128,0.00489777781897121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,256,0.015213333898120456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,64,0.004912888838185204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,8192,32,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,32,0.014137778017255994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,64,0.014813333749771118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,65536,0.3462284406026204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,51200,0.2656986713409424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,65536,0.5336728625827366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,8192,128,0.01479200025399526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,16384,0.08608000146018134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,12288,0.06632355848948161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,51200,0.4232986768086751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,16384,0.13311555650499132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,10240,0.05912444326612684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,8192,0.048525333404541016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,10240,0.08747644556893243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,7168,0.04079022341304355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,8192,0.07214133607016669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,12288,0.10270755820804173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,6144,0.0373946660094791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,5120,0.03168088859981961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,7168,0.06350044409434001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,4096,0.02472266720400916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,6144,0.056299553977118603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,5120,0.04771022333039177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,3584,0.022311111291249592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,3072,0.019285332825448778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,4096,0.03949422306484646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,2560,0.0165928883685006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,3584,0.03605866763326857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,2048,0.014892444014549255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,3072,0.03311199943224589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,2560,0.029675556553734675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,1536,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,1024,0.009383999639087254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,2048,0.02495555579662323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,768,0.007748444047239091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,1536,0.021734222769737244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,512,0.0059262220230367445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,1024,0.017975111802419026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,768,0.016581333345837064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,256,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,128,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,512,0.015557333827018738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,64,0.0041697778635554844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,256,0.01482488877243466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,7168,32,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,128,0.014519110321998596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,64,0.014204444156752693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,7168,32,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,65536,0.257475561565823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,51200,0.20627289348178438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,16384,0.06947466399934557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,65536,0.49140087763468426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,12288,0.0525511105855306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,16384,0.12270932727389866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,10240,0.045590221881866455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,12288,0.09464177820417617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,8192,0.03708088729116652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,10240,0.08084977997673883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,7168,0.03341866532961527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,8192,0.0669048892127143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,6144,0.0288426677385966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,7168,0.058338668611314565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,5120,0.024659555819299486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,6144,0.05108266737726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,4096,0.019650666250122916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,5120,0.043938666582107544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,3584,0.01819644371668498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,4096,0.03680177860789829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,51200,0.38852622773912215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,3072,0.015544000599119397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,3584,0.033991999096340604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,2560,0.013951111170980664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,2048,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,2560,0.028038223584493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,3072,0.031145777967241075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,1536,0.010261333651012844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,2048,0.023580445183648005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,1024,0.007988444632954067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,768,0.0063697778516345555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,1536,0.020351111888885498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,1024,0.017277333471510146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,512,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,768,0.01607288916905721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,256,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,512,0.015142222245534262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,128,0.004335999902751711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,256,0.014491554763582019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,64,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,128,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,6144,32,0.003941333128346337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,64,0.013515555196338229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,6144,32,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,65536,0.24784800741407606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,51200,0.19355555375417074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,65536,0.45005957285563153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,16384,0.06600622336069743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,51200,0.3552230993906657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,12288,0.050937778419918485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,16384,0.11226399739583333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,10240,0.04367466767628988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,12288,0.0870328876707289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,8192,0.03601333167817857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,10240,0.07447733481725057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,7168,0.03153955605294969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,8192,0.06122844749026828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,6144,0.027448000179396734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,7168,0.05316266748640272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,5120,0.023342221975326538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,6144,0.04715199934111702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,4096,0.019268444842762418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,5120,0.04043644335534837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,3584,0.019664888580640156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,4096,0.03451910946104262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,3072,0.016490666402710807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,3584,0.03160711129506429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,2560,0.01533333294921451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,3072,0.028967999749713477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,2048,0.012478222449620565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,2560,0.02633511192268795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,1536,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,2048,0.021781333618693884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,1024,0.008259555531872643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,768,0.006976889239417181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,1536,0.01940444442960951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,512,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,1024,0.016186666157510545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,256,0.004955555415815777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,768,0.015615110596021017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,512,0.014429334137174817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,128,0.0042168887125121225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,64,0.003932444585694207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,256,0.013783999615245394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,5120,32,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,128,0.013516444298956128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,64,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,5120,32,0.012747555143303342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,65536,0.20348800553215873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,51200,0.1598400010002984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,16384,0.05486666825082567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,65536,0.40189777480231387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,51200,0.3195422225528293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,12288,0.04236355423927307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,16384,0.10160089201397365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,10240,0.03600000010596381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,8192,0.029482665989134047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,12288,0.07898400227228801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,10240,0.06740533643298678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,7168,0.02584444483121236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,6144,0.022673777408070032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,8192,0.054401778512530856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,5120,0.01926399932967292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,7168,0.04755022128423055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,4096,0.016183111402723525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,6144,0.04242488741874695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,5120,0.0370142228073544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,4096,0.03158933255407546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,3072,0.013024888932704926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,3584,0.02914311157332526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,2560,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,3072,0.026782222919993933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,2048,0.00981777740849389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,2560,0.02390755547417535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,1536,0.008147555920812819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,2048,0.02012711101108127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,1024,0.006237333433495627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,1536,0.017536888519922893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,768,0.005561777700980504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,1024,0.015501333607567681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,512,0.004886222382386525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,768,0.014759999182489185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,256,0.004171555654870139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,3584,0.01477688882086012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,512,0.013820444544156393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,128,0.004359999878538979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,256,0.013399110900031196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,64,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,4096,32,0.0035128887328836652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,128,0.012864000267452665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,64,0.01276000009642707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,4096,32,0.012736000120639801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,65536,0.17075733343760172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,51200,0.13328444957733154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,65536,0.3952293395996094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,51200,0.309712012608846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,16384,0.04532000091340807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,12288,0.0354364448123508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,16384,0.10004266765382554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,10240,0.030241777499516804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,12288,0.07774755689832899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,8192,0.025003555748197768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,10240,0.06593422094980876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,7168,0.021853332718213398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,8192,0.053540445036358304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,6144,0.01905422243807051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,7168,0.04658133453792996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,5120,0.01643822259373135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,6144,0.04155110981729295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,4096,0.01389422184891171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,5120,0.036736889017952815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,3584,0.01255644443962309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,4096,0.03131822082731459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,3072,0.011230222052998014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,3584,0.028935998678207397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,2560,0.010223111344708337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,3072,0.026566222310066223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,2048,0.008722667064931657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,2560,0.022994667291641235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,1536,0.007624000310897827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,2048,0.019650666250122916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,1024,0.005600888695981767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,1536,0.017518222332000732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,768,0.005221333354711533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,1024,0.01519377695189582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,512,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,768,0.014523555835088095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,256,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,512,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,128,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,256,0.012746666868527731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,64,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,128,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3584,32,0.003495111233658261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,64,0.012418666647540199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3584,32,0.01239911135700014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,65536,0.16191200415293375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,51200,0.12722310754987928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,65536,0.36171555519104004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,16384,0.0437155564626058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,12288,0.033879998657438494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,51200,0.28421955638461643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,10240,0.02905511193805271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,16384,0.09128533469306098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,12288,0.07050844695832995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,8192,0.023451555106374953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,7168,0.02073955535888672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,10240,0.05975466966629028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,8192,0.04828800095452202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,6144,0.018209778600268893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,5120,0.015834665960735746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,7168,0.04320266511705187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,4096,0.013634666800498962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,5120,0.033915556139416166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,3584,0.012126222252845764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,4096,0.029110223054885864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,3584,0.02678488857216305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,3072,0.011344888971911537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,2560,0.009758222434255812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,3072,0.02408355474472046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,2048,0.008649778034951951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,2560,0.022150221798155043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,1536,0.007347555624114142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,2048,0.018626666731304593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,1024,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,6144,0.03833333320087857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,768,0.005223999834722943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,1024,0.014568888478808932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,512,0.00480266660451889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,768,0.0144195556640625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,256,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,256,0.012738666600651212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,128,0.0038720000949170855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,128,0.012523555921183692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,64,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,1536,0.016906667086813185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,64,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,3072,32,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,32,0.012163555456532372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,65536,0.14965689182281494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,3072,512,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,51200,0.11794311470455593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,16384,0.040566222535239324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,65536,0.3551502227783203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,51200,0.27869510650634766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,12288,0.033430222007963396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,10240,0.02872977654139201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,12288,0.06937066713968913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,10240,0.058295110861460366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,8192,0.023223110371165808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,7168,0.020854221449957952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,8192,0.047334220674302846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,6144,0.018364444375038147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,7168,0.042262223031785756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,6144,0.03780533207787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,5120,0.016202666693263583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,5120,0.033407999409569636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,4096,0.013791999883121915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,16384,0.09036978085835774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,3584,0.011818666425016193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,4096,0.028274665276209515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,3072,0.011328889264000786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,3584,0.026426666312747534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,3072,0.023059555225902136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,2560,0.009703999592198266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,2048,0.008020444048775567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,2560,0.021252445048756067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,1536,0.007092444433106317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,2048,0.01885511146651374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,1024,0.005676444619894028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,1536,0.016346666547987197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,768,0.0053288886944452924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,1024,0.014588443769348992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,768,0.014166222678290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,512,0.004443555656406614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,256,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,512,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,128,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,256,0.01309511148267322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,128,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,64,0.0032337777730491427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2560,32,0.003647110942337248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,64,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,65536,0.1361288891898261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2560,32,0.012422222230169507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,51200,0.10761333174175686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,16384,0.03795999950832791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,65536,0.31736000378926593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,12288,0.03249866763750712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,51200,0.2524871031443278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,16384,0.0808159973886278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,10240,0.024400000770886738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,12288,0.06171466906865438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,8192,0.02088088956144121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,7168,0.017796445224020217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,8192,0.04222933451334635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,10240,0.05260533425543043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,6144,0.017704889178276062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,5120,0.01423466702302297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,7168,0.03826399975352817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,6144,0.0346942212846544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,5120,0.03080000148879157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,3584,0.011175110936164856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,4096,0.025441777375009324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,3072,0.009384888741705153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,3584,0.024116444918844435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,2560,0.009146666361225976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,3072,0.021393777595625982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,2048,0.007337777978844113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,2560,0.019626667102177937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,1536,0.006625777731339137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,2048,0.017239999439981248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,1024,0.005246222019195557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,4096,0.011853333148691388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,1024,0.013845333622561561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,768,0.004868444469239977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,512,0.0042568888101312844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,768,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,256,0.0035955554081334006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,512,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,128,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,256,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,64,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,64,0.011939555406570435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,2048,32,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,1536,0.015551111764378019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,128,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,65536,0.09136888715955947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,2048,32,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,51200,0.07271022266811795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,16384,0.02979555394914415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,65536,0.31284621026780873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,51200,0.24771645334031847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,16384,0.07856888903511895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,12288,0.024384000235133704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,10240,0.022129777404997084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,8192,0.01866311166021559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,12288,0.06091377470228407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,7168,0.01581333412064446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,10240,0.05144355694452921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,8192,0.0414959987004598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,6144,0.014113777213626437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,5120,0.012695999609099494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,7168,0.038118223349253334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,4096,0.009005332986513773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,6144,0.03364177876048618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,3584,0.008919110728634728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,4096,0.02473066747188568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,3072,0.007661333514584436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,3584,0.02307644486427307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,5120,0.02962488929430644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,2560,0.007065777977307637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,2048,0.005984000033802456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,3072,0.021313778228229944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,2560,0.0199644449684355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,2048,0.01688799924320645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,1536,0.005330666485759947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,1024,0.004943999979231092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,1536,0.015218666858143277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,768,0.004210666649871402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,1024,0.01385244478782018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,512,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,768,0.013497778111033969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,256,0.003506666670242945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,512,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,256,0.01216533366176817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,128,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,64,0.0033208889265855155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,128,0.01184800018866857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1536,32,0.0035128887328836652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,32,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1536,64,0.01183199965291553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,65536,0.06195377641254001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,51200,0.05003289050526089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,65536,0.2744799984825982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,16384,0.020599111914634705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,51200,0.21630488501654732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,12288,0.017856000198258292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,16384,0.06904355684916179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,12288,0.052299555804994374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,10240,0.019164444671736825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,8192,0.016216889023780823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,10240,0.04433866673045688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,7168,0.013400000002649097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,8192,0.03643644518322415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,7168,0.033253331979115806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,6144,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,5120,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,6144,0.029923554923799302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,4096,0.009083555804358589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,3584,0.008076444268226624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,5120,0.026239110363854304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,4096,0.02239111065864563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,3072,0.008325333396593729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,2560,0.006672888994216919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,3584,0.021018667353524104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,3072,0.01930488811598884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,2048,0.00600888869828648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,1536,0.005305777821275923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,2048,0.015849777393870883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,1024,0.004592888885074192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,1536,0.014535110857751636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,768,0.00416711096962293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,1024,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,512,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,768,0.013081777426931592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,512,0.012359110845459832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,256,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,128,0.0031555555760860443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,256,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,64,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,128,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,1024,32,0.0032053333189752367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,2560,0.018545778261290658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,64,0.01109777804878023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,65536,0.047437333398395114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,51200,0.040716442796919085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,1024,32,0.01147822207874722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,16384,0.0173342227935791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,65536,0.27331911192999947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,12288,0.015603555573357476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,51200,0.215009782049391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,10240,0.0170817772547404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,12288,0.051326221889919706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,16384,0.06814044713973999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,8192,0.015187554889255099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,10240,0.04300533400641548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,7168,0.014041778114106921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,6144,0.01039644413524204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,8192,0.03599822190072801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,7168,0.0325217776828342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,5120,0.00925511121749878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,6144,0.02825066778394911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,5120,0.025430222352345783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,3584,0.007621333003044128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,4096,0.02204711072974735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,3072,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,3584,0.020643555455737643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,2560,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,3072,0.019090667366981506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,2048,0.005956444475385878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,2560,0.01796355512407091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,1536,0.004960000101062986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,2048,0.015565334094895257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,1024,0.004520888957712385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,1536,0.014548444085650973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,768,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,1024,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,4096,0.008007110820876228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,768,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,256,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,512,0.012479110724396177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,128,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,256,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,128,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,64,0.0032631111227803757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,512,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,768,32,0.0031679999083280563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,64,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,65536,0.037258668078316584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,768,32,0.011123555401961008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,51200,0.03122399912940131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,65536,0.2730693287319607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,51200,0.21513688564300537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,16384,0.014927110738224454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,12288,0.013887999786270989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,16384,0.06657155354817708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,10240,0.012452444268597497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,12288,0.05067288875579834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,10240,0.043169776598612465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,8192,0.012533333566453723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,7168,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,8192,0.03535644544495477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,6144,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,7168,0.031693332725101046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,5120,0.008795555267069075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,6144,0.028157333532969158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,4096,0.007678222325113084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,5120,0.025489777326583862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,4096,0.021722666091389124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,3584,0.0070257774657673305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,3072,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,3584,0.020398222737842135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,2560,0.0059075554211934405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,2048,0.005447111195988125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,3072,0.019243554936514962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,2560,0.017620444297790527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,1536,0.004868444469239977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,1024,0.004269333349333869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,1536,0.014202665951516895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,2048,0.015671110815472074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,768,0.003948444293604957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,512,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,1024,0.013201778133710226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,768,0.012896888785892062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,256,0.0032924444725116095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,512,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,128,0.0030826667530669104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,64,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,128,0.01148088855875863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,512,32,0.0028026666906144884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,64,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,32,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,65536,0.025077333052953083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,51200,0.023783110909991797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,65536,0.27221777704026967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,16384,0.01219555570019616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,51200,0.2146231068505181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,12288,0.010374222364690568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,16384,0.06576889091067843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,10240,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,512,256,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,8192,0.010044444766309526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,12288,0.04973777797487047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,10240,0.04152355591456095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,7168,0.009394666386975182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,8192,0.0350888901286655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,7168,0.031094223260879517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,5120,0.00869333330127928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,6144,0.028174221515655518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,4096,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,5120,0.025403555896547105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,3584,0.0069955554273393415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,4096,0.021962665849261816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,3072,0.006275555739800136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,3584,0.02033422225051456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,2560,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,3072,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,2048,0.00499199993080563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,2560,0.01754399968518151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,1536,0.004629333400064045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,2048,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,6144,0.008742222355471717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,1024,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,768,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,1536,0.014175999495718213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,512,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,1024,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,768,0.01276622215906779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,256,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,512,0.012075555821259817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,256,0.011720889144473605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,64,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,128,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,32,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,64,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,65536,0.01829777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,256,32,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,51200,0.01616533266173469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,65536,0.2719564437866211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,51200,0.21357866128285727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,16384,0.009738666315873465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,12288,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,256,128,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,16384,0.06515910890367296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,10240,0.010787555740939246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,12288,0.048708445496029325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,8192,0.009685333404276106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,10240,0.04149422380659316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,7168,0.009342222577995723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,8192,0.034104890293545194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,6144,0.008453333543406593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,7168,0.030893332428402368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,5120,0.00835111074977451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,6144,0.02812533246146308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,4096,0.007314666277832455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,5120,0.02513955533504486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,3584,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,4096,0.021677333447668288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,3072,0.005945777727497949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,3584,0.020243555307388306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,2560,0.0058755555914507965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,3072,0.018886221779717337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,2048,0.00490488898422983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,2560,0.017515555024147034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,1536,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,2048,0.015480000111791821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,1024,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,1536,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,768,0.003847111016511917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,1024,0.012795555922720166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,512,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,768,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,256,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,512,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,128,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,256,0.011384888655609555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,64,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,128,32,0.0026808888134029177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,128,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,65536,0.015206222732861837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,64,0.011040888726711273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,51200,0.015160888433456421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,128,32,0.01071466671095954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,16384,0.009320889082219865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,65536,0.27171466085645885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,51200,0.21326222684648302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,12288,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,10240,0.01034311122364468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,12288,0.04862311151292589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,8192,0.009378666679064432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,10240,0.04115110966894362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,7168,0.008723555339707268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,8192,0.033928887711630926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,7168,0.031175110075208876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,6144,0.028121779362360638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,5120,0.008343111309740279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,5120,0.02535199953450097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,16384,0.06565155585606892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,4096,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,3584,0.006582222051090664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,4096,0.021713778376579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,3072,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,3584,0.020248000820477802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,2560,0.005508444375462002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,3072,0.01863466699918111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,2048,0.004859555512666702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,2560,0.01757333344883389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,6144,0.008691555923885768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,1536,0.0041964443193541635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,1024,0.003855111284388436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,2048,0.01514311134815216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,1536,0.013796444568369122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,768,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,1024,0.012737777498033313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,512,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,256,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,768,0.012198222180207571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,128,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,512,0.012048888537618848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,64,0.002487999904486868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,256,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,64,32,0.0028328889360030494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,128,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,64,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,65536,0.03302133414480422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,64,32,0.01068800025516086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,51200,0.02164799968401591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,16384,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,65536,0.2712222205268012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,51200,0.2131706608666314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,16384,0.06523111131456164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,12288,0.013393777940008374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,10240,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,8192,0.010693333215183683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,12288,0.04829422301716275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,10240,0.04069777660899692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,7168,0.009992889232105678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,8192,0.033725334538353816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,6144,0.009349333743254343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,5120,0.008300444318188561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,7168,0.031234665049446955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,4096,0.006937777830494775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,6144,0.027836445305082533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,5120,0.02511111067401038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,3584,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,4096,0.02169688873820835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,3072,0.005903111149867375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,3584,0.020265777905782063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,2560,0.005381333331267039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,3072,0.018936889039145578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,2048,0.00488800017370118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,2560,0.01756977703836229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,1536,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,2048,0.015233778291278414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,1024,0.0038488888078265717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,1536,0.014112000664075216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,768,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,1024,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,512,0.003273777870668305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,768,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,256,0.002992000016901228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,512,0.012040889097584618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,128,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,256,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,64,0.0026542221506436667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,128,0.011361777782440186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,513,32,32,0.002587555597225825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,64,0.01069155583779017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,513,32,32,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,51200,1.7168977525499132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,16384,0.5744186507331001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,16384,0.6539368629455566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,51200,2.175837410820855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,12288,0.4222417672475179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,12288,0.4846533139546712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,16384,0.5557857619391547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,10240,0.35860710673862034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,12288,0.42529066403706867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,10240,0.4219297832912869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,8192,0.2940231164296468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,8192,0.33702932463751895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,10240,0.35934845606486004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,7168,0.224726226594713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,51200,1.7036915885077581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,8192,0.2943066755930583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,6144,0.21067467000749376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,7168,0.2863413227929009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,6144,0.25086045265197754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,7168,0.2615893416934543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,5120,0.19254488415188262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,6144,0.22921421792772081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,4096,0.14092089070214167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,5120,0.21002044942643908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,4096,0.17204177379608154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,3584,0.13081244627634683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,5120,0.22520799107021758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,3584,0.1513706710603502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,3072,0.12002578046586777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,3072,0.13248444928063288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,3584,0.14649777942233616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,2560,0.08698666758007473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,3072,0.13236355781555176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,2048,0.07983555396397908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,2560,0.1160586675008138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,2560,0.11520889070298936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,2048,0.09556711382336086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,1536,0.05577066871854994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,4096,0.16266932752397326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,1536,0.07688533597522311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,1024,0.045195556349224515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,2048,0.09908088710572983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,1536,0.08287377489937676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,768,0.03501777847607931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,1024,0.06012355619006687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,1024,0.0682248870531718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,512,0.026791999737421673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,768,0.0502800014283922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,256,0.018763555420769584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,512,0.0446284446451399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,512,0.05178755521774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,768,0.06061155266231961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,128,0.01684533390733931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,256,0.039204445150163435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,256,0.04554666744338142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,64,0.01482222146458096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,65536,128,0.03842133283615112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,128,0.037224001354641385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,64,0.03603466682963901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,65536,32,0.03547999925083584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,65536,32,0.015319999721315173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,65536,1.6630311542087133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,65536,2.129464891221788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,51200,1.2992799546983507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,16384,0.43877776463826496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,51200,1.7291751437717016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,16384,0.5210186640421549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,12288,0.3271431128184001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,12288,0.38938222991095645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,16384,0.4554773436652289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,10240,0.27305510309007436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,12288,0.34993865754869247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,10240,0.32384710841708714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,8192,0.22261423534817168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,65536,1.7858462863498266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,10240,0.2964417669508192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,51200,1.3867599699232314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,8192,0.27273066838582355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,7168,0.19736711184183756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,7168,0.22969065772162545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,8192,0.24390043152703178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,6144,0.1667839950985379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,6144,0.2004284461339315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,5120,0.14919643931918675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,6144,0.19080889225006104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,7168,0.2165022161271837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,4096,0.1179271141688029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,5120,0.1696488857269287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,4096,0.13904710610707602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,3584,0.10408888922797309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,5120,0.16485867235395643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,4096,0.1364604499604967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,3072,0.09013688564300537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,3584,0.12328710820939805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,3584,0.12309866481357151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,3072,0.10866489013036092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,2560,0.0746684471766154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,3072,0.11139378282758926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,2560,0.09477244483100043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,2048,0.06368177466922335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,2560,0.098544889026218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,2048,0.07858844598134358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,1536,0.04828977915975782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,2048,0.08422755532794529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,1024,0.036850667662090726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,1536,0.06349688768386841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,1536,0.07127377721998426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,1024,0.050464888413747154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,768,0.030631999174753826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,1024,0.0602160029941135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,512,0.022333333889643352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,768,0.05281155639224582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,768,0.04195466637611389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,256,0.018204443984561496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,512,0.036574221319622464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,512,0.046244445774290294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,128,0.01579466627703773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,256,0.031940443648232356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,256,0.040784888797336154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,64,0.014829332629839579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,51200,128,0.03460089034504361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,128,0.030404445197847154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,51200,32,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,64,0.03019733230272929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,51200,32,0.02921066681543986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,65536,0.551404423183865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,51200,0.42963555124070907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,65536,0.8250026702880859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,51200,0.6601573096381294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,16384,0.1477208932240804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,16384,0.20615643925136992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,65536,0.8558079931471082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,12288,0.11239377657572429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,51200,0.6745057635837131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,16384,0.22236622704399955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,12288,0.15718132919735378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,10240,0.09616977638668484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,10240,0.13219288984934488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,12288,0.17365688747829863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,8192,0.07798400190141466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,10240,0.14803200297885472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,7168,0.07008355855941772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,8192,0.10815644264221191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,8192,0.12356532944573297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,6144,0.061568889353010386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,7168,0.09676977660920884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,7168,0.11100711425145467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,5120,0.05162577827771505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,6144,0.0853911108440823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,6144,0.10002844201193915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,5120,0.07374399900436401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,4096,0.04239911172125074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,5120,0.08755555417802598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,3584,0.03839733203252157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,4096,0.06028888622919718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,4096,0.07427289088567098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,3072,0.03349422083960639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,3584,0.06795999738905165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,2560,0.02896089024013943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,3072,0.04884888728459676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,3072,0.06341333521736993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,2048,0.023717333873112995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,2560,0.04282133446799385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,2560,0.05735466877619425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,3584,0.055348445971806846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,1536,0.019710222880045574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,2048,0.03453422254986233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,2048,0.04970666766166687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,1536,0.043623109658559166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,1024,0.024569777978791132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,768,0.011014222270912595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,1024,0.03811288873354594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,1536,0.029554665088653564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,768,0.02129688858985901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,768,0.03509333398607042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,512,0.019223110543357003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,512,0.03170488940344916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,256,0.0092257774538464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,1024,0.014679110712475248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,256,0.017902221944597032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,128,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,256,0.02795377704832289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,128,0.01739644507567088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,16384,128,0.02572533322705163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,64,0.007335111498832703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,512,0.009715555442704095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,16384,32,0.007910221815109253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,64,0.017256889078352187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,65536,0.4166693422529433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,16384,32,0.017295110556814406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,51200,0.3410328759087457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,51200,0.5408444404602051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,65536,0.6839973131815592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,16384,0.10879821909798516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,65536,0.7243297894795736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,16384,0.17236177126566568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,12288,0.08315733406278822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,16384,0.18995022773742676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,12288,0.13240267170800105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,10240,0.07584089040756226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,51200,0.5712346500820583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,12288,0.1482951111263699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,8192,0.06252088811662462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,10240,0.128257777955797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,8192,0.09168710973527695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,7168,0.05468089050716824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,8192,0.10752177900738186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,10240,0.11252178086174859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,7168,0.09717333316802979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,6144,0.04715822140375773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,6144,0.07256888680987887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,5120,0.03937244415283203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,7168,0.08143999841478136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,5120,0.06406844324535794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,4096,0.032215111785464816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,5120,0.07765066623687744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,4096,0.06569155719545153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,3584,0.029391109943389893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,4096,0.05159999926884969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,6144,0.08690577745437622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,3072,0.025136889682875738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,3584,0.04586755567126804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,3584,0.060639109876420766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,3072,0.04078933265474107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,3072,0.05663466453552246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,2560,0.036791112687852644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,2560,0.051391111479865186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,2048,0.018358222312397428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,2048,0.030233777231640283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,2048,0.04512622290187412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,1536,0.015153777268197803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,1536,0.02602488961484697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,1536,0.03997955719629923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,1024,0.01132266637351778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,1024,0.021393777595625982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,768,0.010021333065297868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,2560,0.021730666359265644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,768,0.019004444281260174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,512,0.008193777667151557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,768,0.03370489014519586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,512,0.01716711123784383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,256,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,256,0.016140444411171805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,512,0.028671112325456407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,128,0.0063697778516345555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,256,0.02714755634466807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,1024,0.03571110963821411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,64,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,12288,128,0.024030221833123103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,128,0.016328889462682936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,12288,32,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,64,0.015806222955385845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,12288,32,0.015779554843902588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,65536,0.37976622581481934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,51200,0.297760009765625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,65536,0.6081040170457628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,16384,0.10407733254962498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,51200,0.4796489079793294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,16384,0.1532871060901218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,12288,0.08037333356009589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,12288,0.11789688799116348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,16384,0.18898488415612114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,10240,0.06217510832680596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,65536,0.7276284429762098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,51200,0.5707448853386773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,8192,0.0521048903465271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,10240,0.1001431081030104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,12288,0.14813421832190618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,7168,0.046748446093665234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,10240,0.12786132759518093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,8192,0.0817448894182841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,8192,0.10675377978218926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,6144,0.039466665850745306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,7168,0.0730480021900601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,7168,0.09643466605080499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,6144,0.0643671088748508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,5120,0.034126222133636475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,6144,0.08653689093059963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,4096,0.028018666638268366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,5120,0.05496622125307719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,5120,0.07698577642440796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,4096,0.04555911156866285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,3584,0.02516888909869724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,4096,0.06512977679570515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,3072,0.021018667353524104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,3584,0.04055466585689121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,3584,0.06037955813937717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,2560,0.01925066610177358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,3072,0.056088891294267446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,3072,0.03674044542842441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,2048,0.016177778442700703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,2560,0.05103022191259596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,2048,0.027888000011444092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,2560,0.03327111072010464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,1536,0.013532444834709167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,2048,0.04454666707250807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,1024,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,1536,0.03956000010172526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,1536,0.024235554867320593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,1024,0.020322667227851022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,1024,0.03270311156908671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,768,0.00869333330127928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,512,0.007191110816266801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,768,0.017973333597183228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,768,0.030200888713200886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,256,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,512,0.0288426677385966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,512,0.016605334149466622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,256,0.015189333094490899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,128,0.005336888962321811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,256,0.026464000344276428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,64,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,128,0.015184000134468079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,10240,32,0.005256888767083486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,10240,128,0.02386933399571313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,64,0.01493866741657257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,10240,32,0.014662222729788886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,65536,0.29892532030741376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,51200,0.21888800462086996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,65536,0.5346346431308323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,51200,0.4227297835879856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,16384,0.07770844300587972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,65536,0.5949066480000814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,51200,0.4691511260138617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,16384,0.13386489285363093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,12288,0.05849777989917331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,16384,0.1567573282453749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,10240,0.04930399854977926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,12288,0.12351377805074056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,12288,0.1031751102871365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,10240,0.08751466539171006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,8192,0.042134222057130605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,10240,0.10669244660271539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,8192,0.07171911001205444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,7168,0.036095112562179565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,8192,0.09112622340520223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,7168,0.06397866540484957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,6144,0.03230488962597317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,7168,0.0823999974462721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,5120,0.02784088916248745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,6144,0.055861334005991616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,6144,0.07550488577948676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,4096,0.023461333579487268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,5120,0.04818844464090136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,5120,0.06656622224383883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,3584,0.021242666575643752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,4096,0.056848002804650195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,4096,0.040692445304658674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,3584,0.03642933236228095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,3072,0.01882933411333296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,3584,0.052593777577082314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,2560,0.016571554872724745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,3072,0.0329751107427809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,3072,0.0503022207154168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,2560,0.029884444342719182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,2560,0.04539111256599426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,2048,0.02515022291077508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,2048,0.04055288765165541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,1536,0.011964444484975604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,1536,0.021356445219781663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,1536,0.03594222333696153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,1024,0.01829866733815935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,1024,0.030505776405334473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,768,0.007831110722488826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,2048,0.014602666099866232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,768,0.01684799955950843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,512,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,768,0.029490666257010564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,512,0.015523556205961438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,256,0.0053280000057485365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,1024,0.009381333159075843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,512,0.027152889304690894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,256,0.014910222755538093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,256,0.025123554799291823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,128,0.004957333207130432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,128,0.014567111929257711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,64,0.0047662220895290375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,8192,128,0.02344088918632931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,8192,32,0.00637511122557852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,64,0.014527110589875115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,65536,0.24710400899251303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,8192,32,0.014188443620999655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,65536,0.49234935972425675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,51200,0.19507822725507948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,65536,0.5951724582248265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,51200,0.3899706734551324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,16384,0.07650666766696505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,16384,0.12415556112925212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,12288,0.05117155445946587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,16384,0.1568195554945204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,12288,0.09539022048314412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,10240,0.043430222405327686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,12288,0.12415733602311875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,51200,0.4695884386698405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,8192,0.03889066643185086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,10240,0.08164355489942762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,10240,0.10713777939478557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,7168,0.03583377930853102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,8192,0.0903973314497206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,8192,0.06753244664933947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,7168,0.059382220109303795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,6144,0.029647111892700195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,7168,0.08238577842712402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,5120,0.02546399997340308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,6144,0.05173155665397644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,6144,0.07464800278345744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,4096,0.02093244426780277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,5120,0.06588977575302124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,4096,0.056622220410241016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,3584,0.019917334119478863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,4096,0.03788888785574172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,3584,0.033999999364217125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,3584,0.05296088920699226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,3072,0.017683555682500202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,3072,0.049838221735424466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,2560,0.015490666031837463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,3072,0.03160888950030009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,2560,0.02825333343611823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,5120,0.04521066612667508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,2048,0.013380444712109037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,2560,0.04641511042912801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,2048,0.023792889383104112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,2048,0.04005333450105455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,1536,0.010779555473062726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,1024,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,1536,0.035678221119774714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,1024,0.030351112286249798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,768,0.007479999628331926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,1024,0.01756800048881107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,768,0.01588000026014116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,512,0.006038222048017714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,768,0.02918400035964118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,512,0.014875556031862894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,256,0.004940444396601783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,512,0.026555554734336004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,256,0.014150222142537435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,128,0.004223111189073986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,1536,0.020475554797384474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,256,0.025191111697090998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,64,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,7168,128,0.02336888843112522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,128,0.013979555832015144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,7168,32,0.004316444612211651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,32,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,65536,0.1946248875723945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,7168,64,0.014051554931534661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,51200,0.17986044618818495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,65536,0.47527556949191624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,51200,0.37485777007208926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,16384,0.05659111340840658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,65536,0.5938053131103516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,51200,0.46859465705023867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,16384,0.12106311321258545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,12288,0.046021332343419395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,16384,0.15577956040700278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,10240,0.03885511226124234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,12288,0.1238408883412679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,12288,0.09432444307539199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,10240,0.08063999811808269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,8192,0.03218400147226121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,10240,0.10653333531485663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,8192,0.06541422340604994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,7168,0.02860355708334181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,8192,0.09076978100670709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,7168,0.05799021985795763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,7168,0.08165511157777575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,6144,0.02488888965712653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,5120,0.021474666065639917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,6144,0.07370844152238634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,6144,0.05057155423694187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,4096,0.016901334126790363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,5120,0.06590044498443604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,5120,0.04373066624005636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,4096,0.03677511215209961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,3584,0.01647555496957567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,4096,0.056306666798061795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,3584,0.052150222990247935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,3072,0.015306666493415833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,3584,0.033922665648990206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,3072,0.030716445710923936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,2560,0.013854222165213691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,3072,0.0499528878264957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,2048,0.011911999848153857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,2560,0.045271111859215625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,2048,0.03981511129273309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,1536,0.00996800015370051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,2048,0.02355555527740055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,1536,0.019549333386951022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,1024,0.007679999702506595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,1536,0.035884443256590105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,1024,0.01662488944000668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,1024,0.030003554291195337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,768,0.006654222392373615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,2560,0.027285334136750963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,768,0.028247998820410833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,512,0.005596444424655702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,768,0.015955555770132277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,256,0.0052266667286554975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,512,0.014499555031458536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,512,0.025830222500695124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,128,0.004600888739029567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,256,0.025048888391918604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,256,0.014191110928853353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,128,0.01348977784315745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,64,0.004309333446953031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,6144,128,0.022375999225510493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,6144,32,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,64,0.01315822203954061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,6144,32,0.01312800000111262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,65536,0.1965351104736328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,51200,0.15429333845774332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,65536,0.4349831210242377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,51200,0.343669335047404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,16384,0.051530665821499295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,65536,0.5946675406561958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,51200,0.4682017962137858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,16384,0.11130400498708089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,12288,0.0407244430647956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,16384,0.15615467230478922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,10240,0.03501777847607931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,12288,0.12328533331553142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,12288,0.08680088652504815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,10240,0.07377155621846516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,8192,0.028198222319285076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,10240,0.10654399792353313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,7168,0.02535377773973677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,8192,0.059359113375345864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,8192,0.09040533171759711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,6144,0.0219217770629459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,7168,0.05244355400403341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,6144,0.07340000073115031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,6144,0.0466648903157976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,5120,0.04066666629579332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,5120,0.0658044417699178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,4096,0.01583377851380242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,7168,0.08127644326951769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,4096,0.05653599898020426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,5120,0.019336000084877014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,3584,0.014848000473446317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,4096,0.03446933295991685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,3072,0.01295644458797243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,3584,0.052052444881863065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,3584,0.03152622116936578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,2560,0.011855111353927188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,3072,0.049053334527545504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,2560,0.025490666429201763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,3072,0.029164443413416546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,2048,0.010215111076831818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,2560,0.044553776582082115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,1536,0.008893333375453949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,2048,0.0381644434399075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,2048,0.021879111727078755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,1536,0.018586667047606576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,1024,0.007081777685218387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,1024,0.015847111741701763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,768,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,1024,0.029542220963372126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,768,0.014896000425020853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,512,0.004908444566859139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,768,0.028607110182444256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,512,0.014128888646761576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,256,0.004586666822433472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,512,0.02647288971477085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,256,0.01349244432316886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,128,0.004040889028045866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,1536,0.0333555539449056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,128,0.01314755529165268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,128,0.022319111559126113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,5120,256,0.024713777833514746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,64,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,5120,32,0.003993777765168084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,64,0.012834666503800286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,5120,32,0.012816000315878125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,65536,0.16966844929589164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,65536,0.39395822419060605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,51200,0.13391733169555664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,65536,0.463997311062283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,51200,0.3097075621287028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,16384,0.04553866717550489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,16384,0.10089777575598823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,51200,0.3661208947499593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,12288,0.03554133243030972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,16384,0.12432266606224908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,12288,0.07829333013958402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,12288,0.09921866655349731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,10240,0.06678311030069987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,10240,0.08632444673114353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,8192,0.023051554958025616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,8192,0.053520000643200345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,8192,0.07342488898171319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,7168,0.020836444364653692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,7168,0.047285334931479565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,6144,0.01846933364868164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,6144,0.04195288817087809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,10240,0.030256887276967365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,5120,0.015469332536061605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,6144,0.06108355522155762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,5120,0.03692266676161025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,4096,0.013067555924256643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,4096,0.031495110856162176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,7168,0.06639555427763197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,4096,0.04766399992836846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,3584,0.012413333687517377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,3584,0.028925332758161757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,3584,0.04453955425156487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,3072,0.01127822200457255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,3072,0.02680355476008521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,3072,0.042487111356523305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,2560,0.0099973330895106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,5120,0.05460000038146973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,2560,0.03927822245491876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,2560,0.023911999331580266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,2048,0.008816888762844933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,1536,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,2048,0.032920890384250216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,2048,0.01995111174053616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,1024,0.005613333235184352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,1536,0.01752622259987725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,1536,0.02951999836497837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,768,0.005255110975768831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,1024,0.015297777122921415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,1024,0.026735999517970618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,512,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,768,0.02574488851759169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,768,0.014452444182501899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,256,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,512,0.01350933313369751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,512,0.02402666707833608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,128,0.0036915557252036203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,256,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,128,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,128,0.02161600026819441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,64,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,4096,32,0.0038488888078265717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,64,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,4096,32,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,65536,0.14984444777170816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,65536,0.36023910840352374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,51200,0.11937333477867974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,4096,256,0.023665777511066858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,65536,0.46230755911933047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,51200,0.28433601061503094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,16384,0.04107644491725498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,16384,0.09175199932522243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,16384,0.12348177697923447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,12288,0.03091999888420105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,51200,0.36602399084303117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,10240,0.02725511127048069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,12288,0.0718640022807651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,12288,0.09909866915808783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,8192,0.021996445126003687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,10240,0.08526844448513454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,8192,0.04870488908555773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,7168,0.019475556082195707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,8192,0.07263911432690091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,7168,0.043465776575936206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,6144,0.01711466742886437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,7168,0.06632355848948161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,6144,0.03907555672857497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,6144,0.06076800160937839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,5120,0.014935111006100973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,5120,0.03421955638461643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,5120,0.0542791121535831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,4096,0.012551999754375882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,10240,0.06056888898213705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,4096,0.029323554701275293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,4096,0.047255999512142606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,3584,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,3072,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,3584,0.02678666677739885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,3584,0.04390755626890394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,2560,0.009328888522254096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,3072,0.04136888848410712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,2048,0.007717333734035492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,2560,0.022312000393867493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,3072,0.025091555383470323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,2560,0.038126223617129855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,1536,0.00666844430896971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,2048,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,1536,0.016507556041081745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,1024,0.005695999910434087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,1536,0.029648890097935993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,2048,0.03379644287957086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,1024,0.014552889598740472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,1024,0.02644888891114129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,768,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,512,0.004705777598751916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,768,0.025450666745503742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,512,0.02372444503837162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,512,0.01367288910680347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,256,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,256,0.02274666726589203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,256,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,128,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3584,128,0.02128977742460039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,128,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,64,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3584,32,0.003597333199448056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,64,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,32,0.012033777932325998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3584,768,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,65536,0.1436542272567749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,51200,0.11347822348276775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,65536,0.35537510448031956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,65536,0.46275912390814883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,51200,0.2816515498691135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,16384,0.03990755478541056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,51200,0.3656071027119954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,12288,0.030513776673210993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,16384,0.12395111719767253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,16384,0.09068710936440362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,10240,0.02587644424703386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,12288,0.07054666678110759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,12288,0.09859022166993882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,8192,0.021563554803530376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,10240,0.059783114327324756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,10240,0.08502755562464397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,7168,0.019127999742825825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,8192,0.04761955473158094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,7168,0.04251022140185038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,6144,0.017142222987280954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,7168,0.06615110900666979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,8192,0.07327999671300252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,5120,0.014569777581426831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,6144,0.038238220744662814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,6144,0.06094933218426175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,4096,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,5120,0.054197334580951266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,5120,0.03403733174006144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,4096,0.028748446040683325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,3584,0.011490667031870948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,4096,0.046867556042141385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,3072,0.010519111322032081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,3584,0.026343999637497797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,3584,0.044525331921047635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,2560,0.009156444834338294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,3072,0.02475555572244856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,3072,0.041823112302356295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,2048,0.008018666671382057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,2560,0.037039111057917275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,2048,0.01824711097611321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,1536,0.0069653333889113525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,2048,0.03254133462905884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,2560,0.022600889205932617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,1024,0.0057955553962124725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,1536,0.029487109846538965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,1536,0.01620444489849938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,768,0.005245333330498801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,1024,0.014501333236694336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,1024,0.02643199927277035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,512,0.004726222405831019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,768,0.013835555149449242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,768,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,512,0.013106666505336761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,512,0.024401777320437964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,256,0.012465777496496836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,256,0.023415111833148535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,128,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,128,0.012418666647540199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,64,0.003830222205983268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,64,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,32,0.0038399998512532976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,3072,32,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,3072,256,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,65536,0.1164062288072374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,3072,128,0.020979555116759405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,65536,0.3538622326321072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,51200,0.09265422158771092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,65536,0.4621093538072374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,16384,0.03253422180811564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,51200,0.2784248987833659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,51200,0.3650195598602295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,16384,0.08994666735331218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,12288,0.024880000286632117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,12288,0.06925333208507962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,16384,0.12323289447360569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,10240,0.020554666717847187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,12288,0.0983768900235494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,10240,0.08486311303244697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,10240,0.05824266539679634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,8192,0.01796888808409373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,8192,0.04692088895373874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,8192,0.07228266530566745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,7168,0.016154666741689045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,7168,0.041894220643573336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,6144,0.014305777019924588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,7168,0.06640088558197021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,6144,0.037773334317737155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,6144,0.060785777038998075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,5120,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,5120,0.05414933297369215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,5120,0.033843557039896645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,4096,0.028657777441872492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,3584,0.010061333576838175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,4096,0.047021332714292735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,3584,0.026234666506449383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,3072,0.009151111046473185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,3584,0.042613334125942655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,3072,0.02407288882467482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,2560,0.007989333735571967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,3072,0.04152266515625848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,2560,0.021187555458810594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,2560,0.03623288869857788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,2048,0.007245333658324347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,2048,0.018393778138690524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,2048,0.03341777788268195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,1536,0.006014222072230444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,1536,0.029579556650585596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,1536,0.016560888952679105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,1024,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,1024,0.014542222023010254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,4096,0.010896888871987661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,1024,0.02680444386270311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,768,0.004958222309748332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,512,0.004523555437723796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,768,0.025464889076020982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,512,0.01314133322901196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,256,0.0038497779104444715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,512,0.02509511179394192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,256,0.02367199957370758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,256,0.01276711126168569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,128,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2560,128,0.021714667479197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,64,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,128,0.012610666453838348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2560,32,0.0036755556033717263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,64,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,768,0.01350044459104538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,65536,0.09188177850511338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2560,32,0.012383999923865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,51200,0.07271289163165622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,65536,0.3147351211971707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,65536,0.45144176483154297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,16384,0.02752711044417487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,51200,0.24936445554097494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,51200,0.3577724562750922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,12288,0.022592888938056097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,16384,0.0802124473783705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,16384,0.12081866794162327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,10240,0.01907111042075687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,12288,0.06162399715847439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,12288,0.09632266892327203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,8192,0.015523556205961438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,10240,0.08294044600592719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,10240,0.05202933483653598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,8192,0.042493333419164024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,7168,0.014275555809338888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,8192,0.0711137784851922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,6144,0.012740444805887012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,7168,0.0383857786655426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,7168,0.06519466638565063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,5120,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,6144,0.03425333234998915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,6144,0.058992001745435924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,4096,0.00963199966483646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,5120,0.05309066507551405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,5120,0.030905776553683813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,3584,0.008976889153321585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,4096,0.0458222230275472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,3584,0.023549333214759827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,3072,0.007992888490358988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,3584,0.04081777731577555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,3072,0.03944089015324911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,3072,0.021432888176706102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,2560,0.007339555356237624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,2560,0.03594844539960226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,2560,0.019574221637513902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,2048,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,2048,0.031939556201299034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,2048,0.01717511150572035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,1536,0.005611555443869696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,1536,0.015652444627549913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,1536,0.02884799904293484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,1024,0.004971555537647671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,1024,0.013507555756303998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,1024,0.02623733381430308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,768,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,768,0.01315111087428199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,4096,0.02602133320437537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,512,0.004201777693298128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,512,0.012736000120639801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,512,0.024196444286240473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,256,0.0038568890757030914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,256,0.022729777627521094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,128,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,256,0.012155555188655853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,128,0.020647111866209242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,128,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,64,0.0034471110751231513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,2048,768,0.0251511103577084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,2048,32,0.003727999826272329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,65536,0.07963199747933282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,64,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,2048,32,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,51200,0.06376622120539348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,65536,0.311172432369656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,51200,0.24512089623345268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,65536,0.45069244172837997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,16384,0.02683999968899621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,16384,0.07866755459043714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,16384,0.12067111333211263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,51200,0.3565359910329183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,12288,0.018181333939234417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,10240,0.0159599996275372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,12288,0.059818665186564125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,12288,0.09553511275185479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,10240,0.0503111117415958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,8192,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,10240,0.08299555381139119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,8192,0.04167644513977898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,7168,0.012706666356987424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,8192,0.07085422012541029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,6144,0.01130488928821352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,7168,0.03740888833999634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,7168,0.06421066655053033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,5120,0.009766221874290043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,6144,0.033257779147889875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,6144,0.05877244472503662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,4096,0.009295110901196798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,5120,0.029884444342719182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,5120,0.0522017776966095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,3584,0.008287111090289222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,4096,0.043507556120554604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,3584,0.022777777579095628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,3072,0.007600888609886169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,3584,0.04033689035309686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,3072,0.020963556236690942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,2560,0.006666666517655055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,2560,0.01923733287387424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,2560,0.035635554128222995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,2048,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,2048,0.01686755485004849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,4096,0.025433777107132807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,1536,0.005207111024194294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,2048,0.031874666611353554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,1536,0.015447111593352424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,1024,0.004567999806669024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,1536,0.02846133377816942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,3072,0.0384880006313324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,768,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,1024,0.026107556290096704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,768,0.01276711126168569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,768,0.025047999289300706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,512,0.0038319999972979226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,512,0.023961777488390606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,256,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,256,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,256,0.02296977738539378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,128,0.003489777859714296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,1024,0.013723555538389417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,128,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,64,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1536,128,0.021338666478792827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1536,32,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,64,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,65536,0.05220444334877861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,512,0.012415111064910889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1536,32,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,51200,0.0432240002685123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,65536,0.2735271188947889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,51200,0.21555199888017443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,16384,0.019189332922299702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,65536,0.4503777821858724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,16384,0.06838133600023058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,51200,0.35631643401251895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,16384,0.11973067124684651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,12288,0.01756177842617035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,10240,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,12288,0.05152088734838697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,10240,0.04378666811519199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,8192,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,10240,0.08226133055157132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,8192,0.036312000619040594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,7168,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,8192,0.07023200061586168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,7168,0.03215288784768846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,7168,0.06364711125691731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,6144,0.01074399964676963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,12288,0.09497866365644668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,6144,0.029354665014478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,5120,0.009701333112186855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,6144,0.058507555060916476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,5120,0.026781333817376032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,4096,0.008424888882372115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,5120,0.050331556134753756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,4096,0.04314133193757799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,3584,0.007273777491516537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,3584,0.020784889658292133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,3072,0.0070088886552386815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,3072,0.019315555691719055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,3072,0.038485334979163274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,4096,0.02202844454182519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,2560,0.006320888797442119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,2560,0.01792977750301361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,2048,0.0053048887186580235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,2560,0.03566844595803155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,3584,0.039839999543295965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,2048,0.015857777661747403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,1536,0.004997333304749595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,2048,0.03228444523281521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,1024,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,1536,0.014485332700941296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,1536,0.02851822310023838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,768,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,1024,0.026100445124838088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,768,0.0127368892232577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,512,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,768,0.02439911166826884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,1024,0.013424889081054263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,256,0.003539555602603488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,512,0.023748444186316595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,512,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,128,0.0032542221662071017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,256,0.022683555881182354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,256,0.011828444070286222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,64,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,128,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,1024,128,0.0199937770764033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,1024,32,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,65536,0.04778222242991129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,64,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,1024,32,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,51200,0.039841777748531766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,65536,0.27248620986938477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,51200,0.21474666065639922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,16384,0.01774488886197408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,65536,0.45091909832424587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,16384,0.06778311067157321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,51200,0.35608887672424316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,12288,0.01551822159025404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,16384,0.11944977442423503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,10240,0.014174222946166992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,12288,0.0510426660378774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,12288,0.09478844536675347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,8192,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,10240,0.08184978034761217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,10240,0.04336533281538221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,8192,0.03596622082922194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,7168,0.011522666447692446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,8192,0.07044621970918444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,7168,0.032292445500691734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,6144,0.01033155537313885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,7168,0.06379199690288968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,5120,0.00905777762333552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,6144,0.02910400099224514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,6144,0.05757688813739353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,4096,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,5120,0.026755554808510676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,5120,0.05006400081846449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,3584,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,4096,0.022078222698635522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,4096,0.04317511121431986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,3584,0.020298666424221463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,3072,0.006701333241330252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,3584,0.03952266772588094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,2560,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,3072,0.019162666466501024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,3072,0.03857600026660495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,2048,0.00554577757914861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,2560,0.01790577835506863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,2560,0.03535111082924737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,1536,0.004960888789759742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,2048,0.03124177787039015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,2048,0.015807999504937064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,1536,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,1024,0.004373333520359463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,1536,0.028224888775083754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,768,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,1024,0.013040000365840064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,1024,0.026241777671708003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,512,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,768,0.024400888217820063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,768,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,256,0.0032595555401510666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,512,0.02338933282428318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,512,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,256,0.011590222517649332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,256,0.022385777698622808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,128,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,768,128,0.01997155613369412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,64,0.0032968889507982465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,128,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,768,32,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,64,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,65536,0.03352800011634827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,768,32,0.011536888778209686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,65536,0.27217422591315377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,51200,0.030301332473754883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,65536,0.45074134402804905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,51200,0.21392888493008086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,16384,0.013936888840463428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,16384,0.06727288828955756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,51200,0.3561724556816949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,12288,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,16384,0.119649781121148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,12288,0.05119377705785963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,12288,0.0949137806892395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,10240,0.011810666984981961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,10240,0.04212799999448988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,8192,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,10240,0.08263999885982938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,7168,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,8192,0.035071998834609985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,8192,0.07004533211390178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,6144,0.010037333601050908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,7168,0.03128000100453695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,7168,0.06345866786109076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,5120,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,6144,0.028864890336990356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,6144,0.056799113750457764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,4096,0.007436444362004598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,5120,0.0502435564994812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,4096,0.022117333279715642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,3584,0.0070471109615431885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,4096,0.04324088825119866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,3584,0.0203457772731781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,3584,0.040496001640955605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,3072,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,3072,0.019215111931165058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,2560,0.0058853332367208265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,5120,0.025661332739724055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,2560,0.017879999346203275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,2560,0.03500977820820279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,2048,0.0052666668262746595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,2048,0.015440000428093804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,2048,0.0318488876024882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,1536,0.004603555632962121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,1536,0.028176890479193792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,1024,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,3072,0.037993778785069786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,1024,0.012852444416946836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,768,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,1024,0.026086222794320848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,768,0.012429333395428128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,512,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,768,0.02509511179394192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,512,0.01182933317290412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,512,0.024393777052561443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,256,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,256,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,128,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,128,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,128,0.02073777715365092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,64,0.011438222395049201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,512,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,32,0.011198221809334226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,512,1536,0.014475555883513557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,65536,0.024544888072543677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,65536,0.2719448937310113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,65536,0.4500720236036513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,51200,0.21340355608198378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,512,256,0.022343110707071092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,16384,0.011037333144081963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,51200,0.35608887672424316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,16384,0.06592266427146064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,12288,0.012583110895421771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,16384,0.11922489272223578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,51200,0.018033777674039204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,10240,0.011185777684052786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,12288,0.049343112442228526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,12288,0.09481600258085464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,8192,0.009717333647939894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,10240,0.08125244246588813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,10240,0.04235644472969902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,8192,0.034367998441060386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,7168,0.009158222211731805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,8192,0.0691031085120307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,6144,0.008799110849698385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,7168,0.031277332040998675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,6144,0.02844266759024726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,6144,0.05580711364746094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,5120,0.008343111309740279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,4096,0.007833778030342525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,5120,0.025495999389224585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,5120,0.04968711071544223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,4096,0.042949332131279834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,3584,0.0070462218589252895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,3584,0.019967110620604623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,3584,0.039741333987977766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,3072,0.00630666646692488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,7168,0.06343111064698961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,3072,0.01887911061445872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,2560,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,3072,0.03819377885924445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,2560,0.017498667041460674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,4096,0.021683555510309007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,2048,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,2560,0.03581777877277798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,2048,0.015495111544926962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,2048,0.031629333893458046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,1536,0.0046675557063685525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,1024,0.003980444537268745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,1536,0.014093332820468478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,1536,0.028160889943440754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,768,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,1024,0.026150221625963848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,1024,0.013032889200581444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,768,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,768,0.024795555406146582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,512,0.011866666376590729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,256,0.0032586666444937387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,512,0.024096889628304377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,256,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,256,0.02275733318593767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,128,0.0028719999310043124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,256,128,0.01995022263791826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,128,0.011511111425028907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,64,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,32,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,64,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,65536,0.015200000670221118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,256,32,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,256,512,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,51200,0.013280889226330651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,65536,0.271083566877577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,65536,0.45044978459676105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,16384,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,51200,0.3560791015625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,16384,0.06519555383258395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,12288,0.00866844422287411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,16384,0.1190604501300388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,12288,0.048710223701265126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,10240,0.00868800034125646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,12288,0.09471466806199814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,10240,0.040511998865339495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,8192,0.007797333101431529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,10240,0.08140089114507039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,51200,0.21404888894822863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,8192,0.034797334008746676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,7168,0.007718222008811102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,8192,0.06760533650716145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,6144,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,7168,0.030973331795798406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,7168,0.06152266926235623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,6144,0.028176890479193792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,5120,0.007844444778230455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,5120,0.02522666586769952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,4096,0.007360888852013483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,5120,0.0502071115705702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,4096,0.021917333205540974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,3584,0.006966222491529252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,4096,0.0432168874475691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,3584,0.02032088902261522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,3072,0.006116444451941385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,3584,0.03997955719629923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,3072,0.018949333164427016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,6144,0.0554986662334866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,2560,0.005683555371231503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,3072,0.038458665211995445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,2048,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,2560,0.017507554756270517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,2048,0.015607999430762397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,2048,0.031239112218221027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,1536,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,1536,0.013811555173661975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,1024,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,1024,0.012497777740160624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,1024,0.025462221768167283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,768,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,2560,0.03496177660094367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,768,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,512,0.003269333392381668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,768,0.025386666258176167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,512,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,512,0.02371911042266422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,1536,0.028837333122889202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,256,0.011594666375054253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,128,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,256,0.022423999177085027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,128,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,64,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,512,128,128,0.02064622276359134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,32,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,64,0.011209777659840055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,128,32,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,65536,0.014550222290886773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,128,256,0.0031271111220121384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,51200,0.01275466630856196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,65536,0.27080710728963214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,16384,0.009348444640636444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,16384,0.06542844242519803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,12288,0.008017777568764156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,12288,0.047800001170900136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,10240,0.008732444710201686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,8192,0.0079297779334916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,10240,0.04162222146987915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,7168,0.007634667058785756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,8192,0.0343484448062049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,6144,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,7168,0.030880000856187608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,51200,0.21222844388749865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,5120,0.007638221813572778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,4096,0.007371555599901412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,6144,0.027790221903059218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,3584,0.0069155556460221606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,5120,0.025070221887694463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,3072,0.006678222368160884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,4096,0.021689777572949726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,2560,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,3584,0.01998133295112186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,2048,0.005984000033802456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,3072,0.018610666195551556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,1536,0.004592000196377437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,2560,0.017533333765135873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,1024,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,2048,0.015245333313941956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,768,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,1536,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,512,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,1024,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,256,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,768,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,128,0.0027840000887711844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,512,0.011503111157152386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,64,0.0027946666296985415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,256,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,64,32,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,128,0.010749333434634738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,64,0.010704889065689512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,65536,0.018539556198649935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,51200,0.015305777390797934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,64,32,0.011037333144081963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,16384,0.008986666798591614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,65536,0.2705155478583442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,16384,0.06487555636299981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,51200,0.21293421586354574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,12288,0.00816888858874639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,10240,0.007615110940403408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,8192,0.007965332931942409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,12288,0.048042668236626514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,10240,0.04007822275161743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,7168,0.007328888608349695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,6144,0.007280000381999546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,8192,0.03393866618474325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,5120,0.00666133314371109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,7168,0.030886222918828327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,4096,0.007304000357786815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,6144,0.02779910961786906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,3584,0.006821333534187741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,5120,0.025039111574490864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,3072,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,4096,0.021538666552967493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,2560,0.006540444576077991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,3584,0.020240000552601285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,2048,0.005990222096443176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,3072,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,1536,0.004509333521127701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,2560,0.01722133325205909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,1024,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,2048,0.015480000111791821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,768,0.00349866681628757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,1536,0.01383733335468504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,512,0.0031591111587153543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,1024,0.012362666428089142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,256,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,768,0.012381333443853589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,128,0.0028320000403457214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,512,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,64,0.0024951110697454875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,256,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,512,32,32,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,128,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,64,0.01069155583779017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,512,32,32,0.010737777584128909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,16384,0.4069937864939372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,16384,0.41835644510057235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,51200,1.201592869228787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,51200,1.3409235212537978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,12288,0.30096178584628636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,10240,0.245432005988227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,12288,0.30914489428202313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,10240,0.25936712159050834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,8192,0.2026817798614502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,7168,0.1735182205835978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,8192,0.2083902226554023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,6144,0.152584883901808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,7168,0.18460088306003145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,5120,0.12828621599409315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,6144,0.1619173288345337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,4096,0.10333422157499526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,5120,0.13688977559407553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,3584,0.09159111314349705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,4096,0.11208266682094997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,3072,0.07921599679523043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,3584,0.0995582209693061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,2560,0.06796533531612821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,3072,0.0865466660923428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,2048,0.05637333128187391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,1536,0.04329955577850342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,2560,0.0759813321961297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,2048,0.06452355782190959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,1024,0.03162044286727905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,1536,0.052483555343416005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,768,0.02569955587387085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,512,0.019440889358520508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,1024,0.042230221960279674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,256,0.012986666626400419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,768,0.036159111393822566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,128,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,512,0.03150399857097202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,256,0.02772355576356252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,64,0.010397333237859938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,128,0.02646933330429925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,65536,32,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,64,0.026033777329656813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,65536,32,0.028216888507207234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,51200,1.008676422966851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,65536,1.2821733686659071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,65536,1.3543973498874242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,16384,0.3397946622636583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,16384,0.3315555519527859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,12288,0.2539048989613851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,12288,0.24851555294460723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,51200,1.0545226203070748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,8192,0.17213155163658989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,10240,0.2087635596593221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,10240,0.20880977312723795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,7168,0.14881689018673366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,8192,0.1672755612267388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,6144,0.13092622492048475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,7168,0.14976621998680964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,5120,0.10992533630794948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,6144,0.12997510698106554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,4096,0.08972266647550796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,5120,0.11107733514573838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,3584,0.07970755630069308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,4096,0.09107555283440484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,3072,0.07005777623918322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,3584,0.08138400316238403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,3072,0.07150133450826009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,2048,0.04980533321698507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,2560,0.06293155749638875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,1536,0.03917510973082648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,2048,0.05368622144063314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,1024,0.028755555550257366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,1536,0.044610665904151074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,768,0.023197333017985027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,1024,0.03604000144534641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,512,0.01718488832314809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,768,0.030607998371124268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,256,0.012059555285506778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,512,0.02608355548646715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,2560,0.05987289216783312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,256,0.024264888630972967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,64,0.00872444444232517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,128,0.02313066687848833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,64,0.022696000006463792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,32,0.009043555292818282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,51200,32,0.022346667117542688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,65536,0.3836035463545058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,51200,128,0.010624888870451184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,65536,0.5062613487243652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,51200,0.30263733863830566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,16384,0.107113778591156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,12288,0.07875377602047391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,51200,0.39596088727315265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,12288,0.1017377773920695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,10240,0.07152177890141805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,16384,0.13419289059109157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,8192,0.06011199951171875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,7168,0.05149866806136238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,10240,0.08707644542058308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,8192,0.07089066505432129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,6144,0.045441776514053345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,7168,0.06350221898820665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,5120,0.039128889640172325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,4096,0.03257333238919576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,6144,0.05706666575537788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,3584,0.029095109966066148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,5120,0.04999555481804741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,4096,0.040608889526791044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,3072,0.02568888829814063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,2560,0.021969777014520433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,3584,0.03774133324623108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,3072,0.033061332172817655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,2048,0.017882666654056974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,2048,0.02573066618707445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,2560,0.03024266825781928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,1536,0.014560888210932413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,1024,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,1536,0.023064000738991633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,768,0.009727111293209923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,1024,0.019689778486887615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,512,0.008034666379292807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,768,0.018567999203999836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,512,0.016532444291644625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,256,0.005960888746711943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,128,0.0052622221410274506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,64,0.004997333304749595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,256,0.015206222732861837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,16384,32,0.00490133340160052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,128,0.01497333414024777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,64,0.014452444182501899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,16384,32,0.014527999692493014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,65536,0.2932479911380344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,51200,0.230475558174981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,65536,0.41317955652872723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,16384,0.08227199978298612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,12288,0.06066133578618368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,16384,0.11135821872287327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,51200,0.3242479960123698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,12288,0.0851751102341546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,8192,0.043659554587470166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,10240,0.07340533203548856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,8192,0.059828446971045605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,7168,0.037951111793518066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,6144,0.03387199838956197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,7168,0.05420533484882779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,5120,0.029176000091764662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,6144,0.04936800069279141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,4096,0.024193776978386775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,5120,0.042862223254309766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,3584,0.021881777379247878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,10240,0.05206222335497538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,4096,0.03527911172972785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,3584,0.031587554348839655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,2560,0.016492444607946608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,3072,0.02905511193805271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,2048,0.01423288881778717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,2560,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,1536,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,2048,0.02325244413481818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,1024,0.010068444742096795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,1536,0.020696888367335003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,768,0.00868000007337994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,1024,0.017909333109855652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,512,0.006879999819729064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,768,0.016755556066830952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,256,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,3072,0.018555555078718398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,128,0.004590222405062782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,512,0.015536889433860779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,256,0.014163555370436775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,64,0.004326222257481681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,128,0.013651555611027611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,12288,32,0.004280888785918553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,64,0.013882666826248169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,12288,32,0.01421688828203413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,65536,0.2699973318311903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,51200,0.21318666140238443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,16384,0.07479733228683472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,65536,0.36922576692369247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,12288,0.05654755565855238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,51200,0.28907820913526744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,10240,0.048698667022917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,16384,0.10044266780217488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,8192,0.04072355561786228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,12288,0.0772071083386739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,7168,0.03524177935388353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,10240,0.06669333246019152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,8192,0.0543431109852261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,6144,0.03174488743146261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,7168,0.04945333467589485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,5120,0.02745955520206028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,6144,0.044509334696663745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,4096,0.02231377859910329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,3584,0.020050666398472257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,5120,0.03849066628350152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,3072,0.017516444126764934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,4096,0.031537777847713895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,2560,0.015588444140222339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,3584,0.02935733397801717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,2048,0.01332800007528729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,3072,0.02625600000222524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,1536,0.012534221841229333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,1536,0.01960266629854838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,1024,0.009750222166379293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,2048,0.021866665946112737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,768,0.008354666332403818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,1024,0.016520000166363187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,512,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,768,0.016217778126398723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,256,0.004942222187916438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,512,0.014726221561431885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,256,0.013807111316257052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,128,0.004265777766704559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,2560,0.02422844370206197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,64,0.004855999930037392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,128,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,10240,32,0.004408888932731417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,64,0.012890666723251343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,10240,32,0.013456000222100152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,65536,0.20393689473470053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,51200,0.15942666265699598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,65536,0.32629066043429905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,16384,0.05827022261089749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,51200,0.2539359993404812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,12288,0.043086220820744835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,16384,0.08986310826407538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,10240,0.0374017788304223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,12288,0.0691831111907959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,8192,0.03131822082731459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,10240,0.05986399783028496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,7168,0.027237332529491846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,8192,0.04854489035076565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,6144,0.024040000306235418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,7168,0.04493244489034017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,5120,0.02068711154990726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,6144,0.0390897790590922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,4096,0.01865244408448537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,5120,0.03418133325046963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,3584,0.01683022247420417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,3584,0.02648799949222141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,3072,0.014838222000333997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,3072,0.02457155618402693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,2560,0.013076444466908773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,2560,0.022580444812774658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,2048,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,1536,0.009562667045328353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,2048,0.020132443971104093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,4096,0.028484443823496502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,1024,0.007592889169851939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,768,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,1536,0.017608889275126986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,1024,0.01516355574131012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,512,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,256,0.004268444246715969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,768,0.015280000037617154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,128,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,512,0.01405244403415256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,64,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,128,0.013065777719020844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,256,0.013599110974205865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,8192,32,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,64,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,8192,32,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,65536,0.22304354773627388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,51200,0.17595110999213326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,65536,0.29105154673258465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,16384,0.061491555637783475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,12288,0.04695644312434726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,51200,0.2266764375898573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,10240,0.039919998910692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,16384,0.08017066452238295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,8192,0.03354933195643955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,12288,0.061679999033610024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,7168,0.02931200133429633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,10240,0.05369155605634054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,6144,0.025555556019147236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,8192,0.04344533218277825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,5120,0.02178222272131178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,7168,0.03949066665437486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,4096,0.017887999614079792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,6144,0.03530933459599813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,3584,0.01613066593805949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,5120,0.030373334884643555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,3072,0.014502222339312235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,4096,0.026152888933817547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,2560,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,3584,0.02462755474779341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,2048,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,3072,0.02237422267595927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,1536,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,2560,0.020949333906173706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,1024,0.007318221860461765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,2048,0.01888088881969452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,768,0.005664888769388199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,1536,0.016661332713233102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,512,0.005227555417352253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,1024,0.014850666125615438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,256,0.004849777867396672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,768,0.01443822185198466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,128,0.004208000169859992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,512,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,64,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,256,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,7168,32,0.004176889028814104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,128,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,64,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,7168,32,0.012792000340090858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,65536,0.1909715599483914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,51200,0.15082755353715685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,16384,0.05244355400403341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,65536,0.281443542904324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,51200,0.21946577231089273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,12288,0.04012711180580987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,16384,0.07864889171388414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,10240,0.034653332498338484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,12288,0.06121777825885349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,8192,0.028559999333487615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,10240,0.052220443884531655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,7168,0.025428444147109985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,8192,0.04225422276390923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,7168,0.03863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,6144,0.034479111433029175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,5120,0.018794667389657762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,5120,0.030140442980660334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,4096,0.015288000305493673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,4096,0.02572533322705163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,3584,0.013978666729397245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,3584,0.024343111448817786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,3072,0.012479999827014076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,6144,0.022430222895410325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,2560,0.011311111350854238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,3072,0.02244088881545597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,2048,0.009939555492666032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,2560,0.02068711154990726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,1536,0.00830577810605367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,2048,0.018096889058748882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,1024,0.006234666539563074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,1536,0.016488000750541687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,768,0.004935111022657818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,1024,0.014608888162506951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,512,0.004552888787455029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,768,0.014446222119861178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,256,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,512,0.013448889056841532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,128,0.004348444441954295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,256,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,64,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,128,0.012447111308574677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,6144,32,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,64,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,6144,32,0.01202044470442666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,65536,0.18018221855163574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,51200,0.1418559948603312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,65536,0.24578399128384062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,51200,0.19292977121141222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,16384,0.06939111153284709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,12288,0.03798844416936239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,10240,0.03291555576854282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,12288,0.05356088942951626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,8192,0.026923555466863845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,8192,0.03716622127426995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,10240,0.04591822293069628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,7168,0.023810666468408372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,6144,0.020605333977275424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,16384,0.049088888698154025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,5120,0.017828444639841717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,6144,0.030947556098302204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,4096,0.014631110760900708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,5120,0.02732444471783108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,7168,0.03435466686884562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,3584,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,3072,0.012268444730175866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,4096,0.023569777607917786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,2560,0.010682666467295753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,3072,0.020575111111005146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,3584,0.02216266592343648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,2048,0.009736888938479954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,2560,0.018943111101786297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,1536,0.008006222546100616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,1024,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,1536,0.015589333242840238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,768,0.0052986666560173035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,1024,0.014091556270917257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,512,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,2048,0.01705777810679542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,256,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,768,0.013797333670987023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,128,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,512,0.013119999733236102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,64,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,128,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,5120,32,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,256,0.013032000097963544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,65536,0.13360444704691568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,32,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,51200,0.10510755247539944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,65536,0.2404391235775418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,16384,0.03798755672242906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,51200,0.18701688448588052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,12288,0.02912799848450555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,16384,0.06807733244366117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,12288,0.051523556311925255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,5120,64,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,10240,0.0251875552866194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,8192,0.020210666788948905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,7168,0.017992888887723286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,10240,0.04388977752791511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,6144,0.015853333804342482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,7168,0.03315911028120253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,5120,0.014130666851997375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,8192,0.03684622380468581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,6144,0.030545777744717065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,4096,0.011870221959220039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,5120,0.027294221851560805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,4096,0.02310400042268965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,3584,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,3072,0.010033778018421596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,3584,0.021707556313938562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,2560,0.0090506664580769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,3072,0.020040000478426617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,2048,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,2560,0.018230222993426852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,1536,0.006713777780532837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,1536,0.01521511044767168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,1024,0.005356444252861871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,2048,0.01641244524055057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,768,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,1024,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,512,0.0042364444169733255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,768,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,512,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,256,0.004322666674852371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,128,0.003495111233658261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,256,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,64,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,128,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,4096,32,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,64,0.012151111331250934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,65536,0.12449067168765598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,4096,32,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,51200,0.09892355733447605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,65536,0.20584711763593885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,16384,0.03776977790726556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,51200,0.16025955147213405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,16384,0.058279110325707324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,12288,0.029432889488008287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,10240,0.025676444172859192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,8192,0.02161955502298143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,12288,0.04455911119778951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,7168,0.018442667192882962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,10240,0.038615110847685076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,8192,0.0317520002524058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,6144,0.016561778055297006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,7168,0.029489778810077246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,5120,0.014801777071423002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,6144,0.02681333323319753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,4096,0.01223111069864697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,5120,0.024678222007221643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,3584,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,3072,0.010500444306267632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,4096,0.02071822186311086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,2560,0.00941777808798684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,3584,0.01995733380317688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,2048,0.008126222425036961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,2560,0.016334222422705758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,3072,0.01846577723821004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,1536,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,2048,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,1024,0.005642666584915585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,768,0.004952888935804367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,1024,0.0135057775510682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,512,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,768,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,256,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,512,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,128,0.00360622215602133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,256,0.012113778127564324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,64,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,128,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3584,32,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,64,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,65536,0.09872355726030137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,32,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,51200,0.0787351131439209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,65536,0.20252533753712973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,16384,0.03284444411595663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3584,1536,0.014565333724021912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,51200,0.15806578265296087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,12288,0.02423644396993849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,10240,0.021016000045670405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,16384,0.056629333231184215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,8192,0.018010665973027546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,12288,0.043624000416861645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,7168,0.01646844380431705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,8192,0.03128533230887519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,10240,0.03747111227777269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,6144,0.014097778333557976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,7168,0.029255112012227375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,5120,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,4096,0.0107351111041175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,5120,0.02400711178779602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,3584,0.01071200023094813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,6144,0.027261333333121404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,4096,0.020062221421135795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,3072,0.009088888764381409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,2560,0.008525333470768398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,3584,0.019138667318556044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,3072,0.017247110605239868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,2048,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,1536,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,2048,0.014870221416155497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,2560,0.016608888904253643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,1024,0.004977777600288391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,1536,0.014185777968830533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,768,0.004548444516128964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,1024,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,768,0.013207111093733044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,512,0.004004444513056013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,256,0.003620444486538569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,512,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,128,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,256,0.012494222157531314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,128,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,64,0.003245333416594399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,3072,32,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,64,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,65536,0.08854933579762776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,3072,32,0.011795555551846823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,51200,0.07193866703245375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,65536,0.19750132825639513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,16384,0.03047200043996175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,51200,0.1544533305697971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,16384,0.05450488792525398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,12288,0.0247999992635515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,10240,0.02443644404411316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,12288,0.042154666450288564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,8192,0.020928889513015747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,10240,0.03697333402103848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,8192,0.03107288810941908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,7168,0.018329777651362948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,6144,0.015930665863884818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,7168,0.028926223516464233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,5120,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,6144,0.02676444417900509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,4096,0.010314666562610203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,5120,0.023100444012218054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,3584,0.009342222577995723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,4096,0.02016088863213857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,3584,0.018328888548745047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,3072,0.008381333616044786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,2560,0.007384888827800751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,3072,0.016943999462657504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,2048,0.00674044465025266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,1536,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,2048,0.014725334114498563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,2560,0.01643822259373135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,1024,0.005281777845488654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,768,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,1536,0.014333332578341166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,512,0.003951111187537511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,768,0.012799999780125089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,1024,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,256,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,128,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,512,0.012519111235936483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,64,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,128,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,256,0.012171555724408893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2560,32,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,64,0.01146133326821857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,65536,0.0693733361032274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,51200,0.055890666113959424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2560,32,0.012077333198653327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,16384,0.02419288953145345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,65536,0.1963822179370456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,12288,0.019127110640207928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,51200,0.15213333235846624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,10240,0.017159110969967313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,12288,0.04112799962361654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,16384,0.053525331947538585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,8192,0.017331555485725403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,7168,0.015581332974963717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,8192,0.030958222018347845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,6144,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,6144,0.026201777988009985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,5120,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,5120,0.023375110493765935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,4096,0.008773333496517604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,10240,0.0364951127105289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,3584,0.00906755526860555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,7168,0.02846933404604594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,4096,0.01963733302222358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,3584,0.018213333355055917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,2560,0.006656000183688269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,2560,0.015882665912310284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,3072,0.017229333519935608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,2048,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,1536,0.005321777943107817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,1536,0.013902222116788229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,2048,0.01482222146458096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,1024,0.004558222161398994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,768,0.004210666649871402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,1024,0.013178666432698568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,768,0.012833777401182385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,512,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,256,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,3072,0.007640000018808577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,512,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,128,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,64,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,256,0.012208888928095499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,128,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,2048,32,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,64,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,2048,32,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,65536,0.055528889099756874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,51200,0.04513066675927904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,65536,0.15769422054290771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,16384,0.01910666624704997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,51200,0.12291467189788818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,16384,0.04320711228582594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,12288,0.016166221764352586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,10240,0.016563556260532804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,12288,0.0341360006067488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,8192,0.015117333994971381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,10240,0.030842665168974135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,7168,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,7168,0.02439555525779724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,6144,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,6144,0.022243556049134996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,5120,0.011311111350854238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,8192,0.025763556361198425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,5120,0.020104888412687514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,4096,0.009583111438486312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,3584,0.009340444372759925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,4096,0.017275555266274344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,3584,0.01615111033121745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,3072,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,2560,0.006769777586062749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,2048,0.005406222409672207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,3072,0.015142222245534262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,1536,0.005015111217896144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,2560,0.014803555276658801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,1024,0.004280888785918553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,2048,0.013611555927329592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,1536,0.013056889176368713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,768,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,512,0.003646222253640493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,768,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,1024,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,256,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,128,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,512,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,64,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,256,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1536,32,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,64,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,65536,0.0407591097884708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,32,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,51200,0.03377155462900797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,65536,0.15602133009168836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,16384,0.014800889624489678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,16384,0.042103999190860324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,51200,0.12035911613040501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,12288,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,10240,0.012671999633312225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1536,128,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,12288,0.03466755482885573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,8192,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,10240,0.02985066506597731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,7168,0.011506666739781698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,6144,0.010508444574144153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,7168,0.022666666242811415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,8192,0.024911110599835713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,5120,0.008989333278603023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,6144,0.02088622252146403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,4096,0.007661333514584436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,3584,0.007290666302045186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,4096,0.01657511128319634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,5120,0.019549333386951022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,3072,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,3584,0.016192000773217943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,2560,0.005986666513813867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,2048,0.0053084443012873335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,3072,0.015447111593352424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,2048,0.013165333204799227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,1536,0.004927999857399198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,1024,0.004230222354332606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,1536,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,768,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,1024,0.012451555165979596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,768,0.012139555480745105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,512,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,256,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,512,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,128,0.0031475555151700974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,256,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,64,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,128,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,1024,32,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,2560,0.014493332968817817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,64,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,65536,0.032587554719712995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,51200,0.028428445259730022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,1024,32,0.011110222174061669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,16384,0.01367111090156767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,51200,0.1190355618794759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,65536,0.15480621655782065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,16384,0.04102399945259094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,12288,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,10240,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,12288,0.034345779154035784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,10240,0.02860266632503933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,8192,0.012519111235936483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,7168,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,8192,0.02473066747188568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,6144,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,7168,0.022667555345429316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,6144,0.02065333392884996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,5120,0.008719111482302347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,4096,0.0074746666683091064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,5120,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,3584,0.007248000138335758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,4096,0.01684799955950843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,3584,0.01590666671593984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,3072,0.014920888675583733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,2560,0.005954666684071223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,2560,0.014163555370436775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,2048,0.005344888816277186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,2048,0.013184000220563678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,1536,0.0046302220887608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,3072,0.006324444380071428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,1536,0.012829333543777466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,1024,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,768,0.003946666502290302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,1024,0.012130666938092975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,512,0.003539555602603488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,768,0.012558221817016602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,256,0.0033902221669753394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,512,0.012004444168673621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,128,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,64,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,256,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,768,32,0.0029644444584846497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,32,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,65536,0.025472000241279602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,65536,0.15461777316199407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,51200,0.02082577844460805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,128,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,768,64,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,16384,0.014160000615649752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,12288,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,51200,0.11868177519904242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,16384,0.040758222341537476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,10240,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,12288,0.0324888891643948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,8192,0.010413332945770687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,7168,0.009389333426952362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,10240,0.02810488806830512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,8192,0.02407466702991062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,6144,0.009742221898502773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,5120,0.008709333009190029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,7168,0.022686221533351477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,6144,0.0206631107462777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,4096,0.007625777688291337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,5120,0.01903288894229465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,3584,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,3072,0.0064088888466358185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,4096,0.016521778371598985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,2560,0.005951111101441913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,3584,0.015853333804342482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,2048,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,3072,0.015211555692884656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,2560,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,1536,0.004604444321658877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,2048,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,1024,0.004212444441186057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,768,0.0038933331767717996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,1536,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,512,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,1024,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,768,0.012515555653307172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,512,0.011853333148691388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,128,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,64,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,512,32,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,128,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,65536,0.018544000056054857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,32,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,51200,0.016190222567982145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,65536,0.1513555579715305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,16384,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,51200,0.11606132984161377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,16384,0.03946222199334039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,12288,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,10240,0.011155555645624796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,12288,0.03137422270245022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,8192,0.009669333696365356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,10240,0.028386665715111628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,512,64,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,8192,0.02405599918630388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,7168,0.009081777599122789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,6144,0.008639111287064022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,5120,0.00832266691658232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,7168,0.022015111313925848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,4096,0.007346666521496243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,6144,0.020651555723614164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,5120,0.01889155473974016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,3584,0.0069511110583941145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,4096,0.016421332955360413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,3072,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,3584,0.01587377819750044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,2560,0.005864000154866113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,3072,0.014837332897716098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,2048,0.004996444616052839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,2560,0.014504000544548035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,2048,0.013204444613721637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,1536,0.004564444637960858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,1024,0.004078222231732475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,1536,0.012796444197495779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,768,0.0038035557501845886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,1024,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,768,0.01238933371173011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,512,0.011832888755533429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,256,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,256,0.011251555548773872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,128,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,128,0.011019555230935415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,64,0.00254133323000537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,64,0.010766222245163388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,32,0.0029279999434947968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,256,32,0.010810666614108615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,65536,0.015287111202875773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,65536,0.1520328919092814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,51200,0.01584533353646596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,16384,0.009397333694828881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,51200,0.11596888966030544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,256,512,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,12288,0.009331555830107795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,10240,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,16384,0.039064887497160174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,8192,0.00942666663063897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,12288,0.03241689006487528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,10240,0.027863999207814533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,7168,0.00904977818330129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,8192,0.023843554986847773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,6144,0.008391111261314815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,7168,0.02180355621708764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,5120,0.00848711116446389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,4096,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,6144,0.020632000433074105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,3584,0.006682666639486949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,5120,0.018591110905011494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,3072,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,4096,0.016584888100624084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,3072,0.014845333165592618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,2560,0.01391999920209249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,2048,0.004872000051869286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,2048,0.012834666503800286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,1536,0.004567999806669024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,1536,0.012486221889654795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,1024,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,3584,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,768,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,1024,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,2560,0.005630222045713001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,512,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,768,0.012503999802801343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,128,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,512,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,64,0.0029724445194005966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,128,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,128,32,0.002604444407754474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,64,0.011508444117175208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,65536,0.023407111565272015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,32,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,51200,0.01889333294497596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,51200,0.11592977576785618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,16384,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,128,256,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,16384,0.03931377662552728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,12288,0.009399111072222391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,10240,0.008393777741326226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,12288,0.031268444326188825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,10240,0.027786665492587622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,8192,0.008055110772450766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,7168,0.00869511150651508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,8192,0.023710222707854375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,6144,0.008012444608741337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,7168,0.02237155536810557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,5120,0.00832622249921163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,65536,0.15201600392659506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,6144,0.020272889071040683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,4096,0.007328000333574083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,5120,0.018921777606010437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,3584,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,3072,0.0062773335311147906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,4096,0.016191111670600045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,2560,0.0053022222386466126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,3584,0.015837333268589444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,2048,0.004914666629499859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,2560,0.013873777455753751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,1536,0.004202666795916027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,2048,0.01331911070479287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,1024,0.00360622215602133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,1536,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,768,0.0035964445107513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,1024,0.012180444267061023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,768,0.012168000141779581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,512,0.003330666571855545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,512,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,256,0.0029537777105967202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,3072,0.015211555692884656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,256,0.011696889168686338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,128,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,64,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,128,0.011085333095656501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,64,32,0.002584888910253843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,64,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,65536,0.02714311083157857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,64,32,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,51200,0.02274755636850993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,65536,0.15131110615200466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,51200,0.11497599548763698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,16384,0.010407110883129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,12288,0.010440000229411654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,16384,0.03839200072818332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,10240,0.009357333183288574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,12288,0.031599998474121094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,8192,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,10240,0.028187556399239436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,7168,0.00871022211180793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,8192,0.0240657776594162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,7168,0.022035555707083807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,6144,0.008345777789751688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,5120,0.008441777692900764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,4096,0.007317333585686154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,6144,0.020647111866209242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,3584,0.006606222026877933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,5120,0.019167999426523846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,3072,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,3584,0.01587199999226464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,2560,0.00526577772365676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,3072,0.015209777487648858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,2048,0.004980444494220945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,2560,0.01421688828203413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,1536,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,2048,0.013437333206335703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,1536,0.012897777888509961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,1024,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,768,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,1024,0.01221777747074763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,512,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,768,0.012447111308574677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,256,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,4096,0.01644622286160787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,128,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,512,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,64,0.002522666628162066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,256,0.011552888486120435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,257,32,32,0.0025635556214385563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,128,0.010766222245163388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,64,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,257,32,32,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,51200,1.0806781980726454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,51200,1.3659573660956488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,16384,0.45304711659749347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,16384,0.3468008836110433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,12288,0.3410808775160048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,12288,0.2616639931996663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,16384,0.2985911104414198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,10240,0.24880978796217176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,12288,0.23046665721469453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,10240,0.22035111321343315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,8192,0.20407733652326795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,8192,0.17720088693830702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,10240,0.19887822204165986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,7168,0.17605600092146131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,51200,0.8797475496927897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,8192,0.16069422827826604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,7168,0.15727910730573866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,6144,0.15451733271280924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,6144,0.13794133398267958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,5120,0.12908177905612522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,5120,0.11738577153947617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,7168,0.1438924471537272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,4096,0.10395999749501546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,6144,0.1276657713784112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,3584,0.09044800202051799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,4096,0.09631288713879055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,3584,0.08529511425230239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,4096,0.09493333101272583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,3072,0.07862844732072619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,3584,0.0859866672092014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,3072,0.07534399959776136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,2560,0.06703644328647189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,2560,0.06649066342247857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,3072,0.0779013302591112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,2048,0.055629332860310875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,2048,0.05654222435421414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,2560,0.07041866911782159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,1536,0.04011999898486667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,2048,0.06227999925613403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,1024,0.02843377656406826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,1536,0.04675111174583435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,1536,0.05466844307051765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,768,0.023359111613697473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,1024,0.03775644302368164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,1024,0.046922667158974536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,768,0.03220444586541917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,512,0.017128888103697035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,768,0.04353866643375821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,512,0.027434666951497395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,512,0.038513776328828596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,256,0.013231111069520315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,5120,0.11183644665612115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,256,0.03501155641343858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,128,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,256,0.02494311167134179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,64,0.010788444843557147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,65536,128,0.029076443778143987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,128,0.023599111371570166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,65536,32,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,64,0.02308355602953169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,65536,32,0.02239822182390425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,65536,1.2647991180419922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,65536,1.0979173448350694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,51200,0.9921929041544596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,51200,0.8664515813191732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,16384,0.3362409008873834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,16384,0.281240012910631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,16384,0.261681768629286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,12288,0.24874401092529297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,12288,0.21273599730597603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,65536,0.9777866999308268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,10240,0.1909324460559421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,10240,0.1787644492255317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,51200,0.7678444650438098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,12288,0.20395821995205352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,8192,0.15839466783735487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,8192,0.1443493366241455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,10240,0.17302933004167345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,7168,0.13463911745283338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,7168,0.12869955433739558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,6144,0.11962933010525173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,8192,0.1416506634818183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,6144,0.11247377925448948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,7168,0.12735111183590359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,5120,0.10004711151123047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,5120,0.09684711032443577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,6144,0.11397955152723525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,4096,0.08102489180035062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,5120,0.09976000256008571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,4096,0.07906488577524821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,3584,0.07205955849753486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,3584,0.07074577940834893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,4096,0.08544355630874634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,3072,0.06269955635070801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,3584,0.07690755526224773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,2560,0.05408622158898247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,3072,0.06330221891403198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,3072,0.07060888740751478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,2048,0.044919110006756253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,2560,0.055627557966444224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,2048,0.04755377769470215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,1536,0.03472355670399136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,2560,0.06468711296717326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,2048,0.05664889017740885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,1024,0.02567200031545427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,1536,0.03980088896221585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,768,0.020815110868877836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,1536,0.0508266654279497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,1024,0.03222666515244378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,512,0.015102222561836243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,1024,0.04399822155634562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,768,0.026966222458415564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,256,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,768,0.04056088791953193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,512,0.0366213321685791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,128,0.00941777808798684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,256,0.021935999393463135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,256,0.0336737765206231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,51200,128,0.02739022175470988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,128,0.020920889245139226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,32,0.008731555607583787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,64,0.02032444377740224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,32,0.020297777321603563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,65536,0.37712711758083767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,51200,512,0.02328088879585266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,65536,0.4342302216423883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,51200,0.2990479999118381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,51200,64,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,16384,0.10343022478951348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,51200,0.34166932106018066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,16384,0.11678932772742377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,12288,0.07776711384455363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,16384,0.12440889411502415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,12288,0.08885155783759223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,10240,0.06632977724075317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,12288,0.09984622399012248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,51200,0.3414151138729519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,65536,0.43671377499898273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,10240,0.07626933521694608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,10240,0.08757333623038398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,8192,0.06167911158667671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,7168,0.04723377691374885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,8192,0.07447733481725057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,7168,0.05563555823432075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,6144,0.04203022188610501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,6144,0.04985688792334663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,6144,0.06287644306818645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,8192,0.056199113527933754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,5120,0.03587466809484694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,5120,0.044844445255067616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,5120,0.05738133192062378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,4096,0.029839999145931665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,4096,0.03595733311441209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,7168,0.06775289111667208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,3584,0.02696977721320258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,4096,0.0498524440659417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,3072,0.023982221881548565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,3584,0.03321511215633816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,3584,0.04583377639452616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,2560,0.020452444752057392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,3072,0.030472887886895075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,3072,0.04344266653060913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,2048,0.016893333858913846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,2560,0.02606577840116289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,2560,0.040511998865339495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,1536,0.013704000247849358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,2048,0.023440000083711412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,2048,0.037067555718951754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,1024,0.010768888725174798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,1536,0.020603555772039626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,1536,0.034641779131359525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,768,0.00921688891119427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,1024,0.018537777993414137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,1024,0.030476444297366675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,512,0.007625777688291337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,768,0.016833777228991192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,768,0.029116445117526587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,512,0.015815999772813585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,512,0.02665333284272088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,256,0.01443733274936676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,128,0.004579555657174853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,256,0.025087111526065405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,128,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,16384,128,0.02130844526820713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,64,0.014124444789356656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,32,0.006231110956933763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,16384,32,0.01586577792962392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,256,0.005570666657553778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,65536,0.2643013265397814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,65536,0.3728497823079427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,16384,64,0.004631999880075455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,51200,0.20831644535064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,51200,0.291470209757487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,16384,0.07622400257322523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,16384,0.10158844126595391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,65536,0.4345048798455133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,12288,0.055381334490246244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,51200,0.338556448618571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,12288,0.07826844188902113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,10240,0.04779111014472114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,12288,0.09920177857081096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,10240,0.06728266345130073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,8192,0.03998400105370416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,10240,0.08726755777994792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,8192,0.055310222837660045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,7168,0.034550223085615374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,16384,0.12402044402228461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,8192,0.07333511114120483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,6144,0.03101244568824768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,7168,0.0499875545501709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,7168,0.06749600172042847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,5120,0.02678933408525255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,6144,0.04587999979654948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,6144,0.0625964469379849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,4096,0.022135110365019903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,5120,0.05659733215967814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,5120,0.039408889081743025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,3584,0.01998666591114468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,4096,0.03241599930657281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,4096,0.04896000027656555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,3584,0.029654221402274236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,3072,0.017416000366210938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,3584,0.045722666713926524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,2560,0.014882667197121514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,3072,0.026742221580611333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,3072,0.0425582230091095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,2048,0.013106666505336761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,2560,0.024119110571013555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,2560,0.039819555150138006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,1536,0.010854221880435944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,2048,0.02132533325089349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,2048,0.03669688767857022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,1024,0.008728889127572378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,1536,0.019341333044899833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,1536,0.03362222181426154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,768,0.00797155582242542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,1024,0.01716711123784383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,1024,0.029804444975323145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,512,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,768,0.0151591118839052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,768,0.027822222974565294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,256,0.004895110925038655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,512,0.014108444253603617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,512,0.025809778107537165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,256,0.013765333427323235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,128,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,256,0.024337778488794964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,64,0.0045422220395671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,128,0.013426666458447775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,12288,128,0.020652444826232064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,12288,32,0.004573333180612988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,32,0.014536889062987434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,12288,64,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,65536,0.22732978396945527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,51200,0.17851111623975965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,65536,0.3314693239000109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,51200,0.25908978780110675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,16384,0.06562399864196777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,16384,0.09126755264070298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,65536,0.4317297670576308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,51200,0.3384737703535292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,12288,0.048482666412989296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,16384,0.12318755520714654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,10240,0.04180799921353658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,12288,0.07048977745903863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,12288,0.09946399927139282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,10240,0.06093244420157539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,8192,0.035402665535608925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,10240,0.08656089173422919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,7168,0.030402666992611353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,8192,0.05011733373006185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,8192,0.07321866353352864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,7168,0.04493955439991421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,6144,0.02735111117362976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,7168,0.06704977485868666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,5120,0.023277334041065637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,6144,0.04094311263826158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,6144,0.062262223826514355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,5120,0.035515556732813515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,5120,0.056045333544413246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,4096,0.019419555862744648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,3584,0.01760088900725047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,4096,0.028959999481836956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,4096,0.04865511258443197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,3072,0.014873777826627096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,3584,0.044998221927218966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,3072,0.024655999408827886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,2560,0.013360888593726687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,3072,0.042280889219707914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,2560,0.022587555978033278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,2560,0.03970222340689765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,2048,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,2048,0.020251555575264823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,1536,0.009730666875839233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,1536,0.01829333272245195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,3584,0.027419555518362258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,1024,0.008012444608741337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,1024,0.016211556063758004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,2048,0.03632355398601956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,1024,0.02995022137959798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,768,0.0068400001360310455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,768,0.015222221612930298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,768,0.02608799934387207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,512,0.005568888866239124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,512,0.01384888920519087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,1536,0.03342933456103007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,256,0.004598222259018156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,512,0.025735111700163946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,128,0.004304000073009067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,256,0.01316088851955202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,128,0.021744888689782884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,64,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,128,0.01314399970902337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,10240,32,0.003974222060706881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,64,0.012801777985360889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,10240,32,0.013373333546850415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,65536,0.20143555270300972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,65536,0.2881813314225939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,10240,256,0.02419288953145345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,51200,0.15840800603230795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,51200,0.22146844863891602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,16384,0.056984000735812716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,65536,0.3029910988277859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,16384,0.08035111427307129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,12288,0.0429004430770874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,12288,0.0625235570801629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,16384,0.09385955333709717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,51200,0.23922401004367402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,10240,0.03681244452794393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,12288,0.07707822322845459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,8192,0.029209779368506536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,10240,0.054661333560943604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,8192,0.043967998690075345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,7168,0.025660443637106154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,8192,0.057335999276902944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,7168,0.04010844561788771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,6144,0.02293599976433648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,7168,0.05283022258016798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,6144,0.03578755590650771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,5120,0.019578667150603402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,6144,0.04983555608325534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,10240,0.06745422548717923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,5120,0.0306640002462599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,4096,0.015556444724400839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,5120,0.04586577746603224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,4096,0.02609066665172577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,3584,0.014017777310477363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,4096,0.041054222318861223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,3584,0.024438222249348957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,3584,0.03821422325240241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,3072,0.012652444342772165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,2560,0.011613333390818702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,3072,0.03562577896647983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,3072,0.02308711078431871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,2560,0.020954666866196528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,2048,0.01035911093155543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,2560,0.033955554167429604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,1536,0.008365333080291748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,2048,0.032009777095582746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,2048,0.01920088960064782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,1024,0.006934222247865465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,1536,0.016886222693655226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,1024,0.025443555580245122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,1536,0.02918133470747206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,768,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,768,0.025380444195535447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,512,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,768,0.014447111222479077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,512,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,512,0.023737778266270954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,256,0.004218666503826777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,256,0.023379556006855432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,128,0.003939555750952827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,256,0.013106666505336761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,128,0.012478222449620565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,1024,0.014816888504558139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,64,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,8192,32,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,64,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,8192,32,0.013767999907334646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,65536,0.17886932690938315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,65536,0.2534417841169569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,51200,0.1401040024227566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,51200,0.19792711734771729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,65536,0.2974880006578234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,8192,128,0.021451556020312842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,16384,0.05077244506941902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,16384,0.06976266702016194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,16384,0.09255733092625935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,12288,0.03801688883039687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,12288,0.05571733580695259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,51200,0.2349635495079888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,10240,0.033399111694759794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,12288,0.07580266396204631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,8192,0.027506666051016912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,10240,0.06594400273429023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,10240,0.048356443643569946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,7168,0.023504889673656885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,8192,0.05588710970348782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,7168,0.035828444692823626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,6144,0.020647111866209242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,7168,0.05148799882994758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,6144,0.031587554348839655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,6144,0.04857866631613838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,5120,0.017437333861986797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,5120,0.02810844447877672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,5120,0.044758223825030856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,4096,0.014373333917723762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,4096,0.024062222904629175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,4096,0.03958488835228814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,3584,0.013214222258991666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,3584,0.022494221727053326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,3584,0.036601778533723615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,8192,0.039453334278530545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,3072,0.011839110818174152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,2560,0.010771555205186209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,3072,0.021104888783560857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,3072,0.03581422236230638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,2560,0.019659555620617337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,2048,0.009492444495360056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,2560,0.033372445238961115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,1536,0.008023999631404877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,2048,0.017627555463049147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,1536,0.015678221980730694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,1536,0.027073777384228174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,1024,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,2048,0.031798223654429116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,768,0.005485333502292633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,1024,0.01421688828203413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,1024,0.02516088883082072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,512,0.004305777864323722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,768,0.024958221448792353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,512,0.013154666456911298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,768,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,256,0.003826666623353958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,512,0.023361777265866596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,128,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,256,0.023759111762046814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,256,0.012839111189047495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,128,0.012110222544935016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,64,0.003432000055909157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,7168,128,0.020958221620983545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,7168,32,0.0036106668412685394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,64,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,7168,32,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,65536,0.16511910491519505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,65536,0.2473564412858751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,51200,0.12804799609714085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,65536,0.29441955354478627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,51200,0.1930097738901774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,16384,0.04682933290799459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,16384,0.07019733058081733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,51200,0.232152885860867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,16384,0.09124977721108331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,12288,0.03514666689766778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,12288,0.054317331976360746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,12288,0.07672888702816434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,10240,0.030679110023710463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,10240,0.04753511150677999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,10240,0.06636533472273085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,8192,0.02548888822396596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,8192,0.055268443293041654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,8192,0.03931555483076308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,7168,0.02421244482199351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,7168,0.03662755423121982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,6144,0.019357333580652874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,7168,0.05398222141795688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,6144,0.032046222024493747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,6144,0.04848177896605598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,5120,0.017673777209387887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,4096,0.01402844488620758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,5120,0.028008000718222723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,5120,0.04460533459981283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,3584,0.012490666574902005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,4096,0.03905955619282193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,4096,0.023727110690540735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,3072,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,3584,0.02235466738541921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,3584,0.036757333411110774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,2560,0.010054222411579555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,3072,0.03492710987726847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,2560,0.019297776950730216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,2560,0.0331857767370012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,2048,0.008735111190213097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,2048,0.017300445172521803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,2048,0.029564443561765883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,1536,0.007663999994595845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,1536,0.015488889482286243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,1536,0.02721244427892897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,1024,0.006367110957702001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,1024,0.014119999276267158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,1024,0.0247688889503479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,768,0.0053075556125905775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,3072,0.020613332589467365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,512,0.004889777965015835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,768,0.013507555756303998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,768,0.02472266720400916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,256,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,512,0.012872888810104795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,512,0.023435556226306494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,128,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,256,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,256,0.022798221972253587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,64,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,6144,128,0.020319999919997323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,128,0.012083555261294046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,6144,32,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,64,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,6144,32,0.013201778133710226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,65536,0.24240978558858237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,51200,0.10461777448654175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,65536,0.29259199566311306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,51200,0.18861599763234457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,16384,0.03776622149679396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,65536,0.1357297764884101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,16384,0.06886577606201172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,12288,0.02864266766442193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,16384,0.08982488844129775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,51200,0.229784886042277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,10240,0.02481777800454034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,12288,0.05272888806131151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,12288,0.07304977708392672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,8192,0.021241777473025854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,10240,0.044799112611346774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,10240,0.0644213358561198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,7168,0.01833511061138577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,8192,0.0372648901409573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,8192,0.05489244394832187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,6144,0.016189333465364244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,7168,0.03323910964859857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,7168,0.05070844292640686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,5120,0.014200000299347771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,6144,0.030878222650951807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,6144,0.0480640000767178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,4096,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,5120,0.027245332797368366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,5120,0.0440959996647305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,3584,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,4096,0.023269333773189124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,4096,0.03926488757133484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,3072,0.009742221898502773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,3584,0.021898667017618816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,3584,0.036201778385374285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,3072,0.02085688875781165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,3072,0.03487822082307603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,2560,0.009045333498054082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,2560,0.03284622232119242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,2560,0.01888177792231242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,2048,0.007987555530336168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,2048,0.016394666499561734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,2048,0.028905779123306274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,1536,0.00700800038046307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,1536,0.014855111638704935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,1536,0.02718577782313029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,1024,0.005681777579916849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,1024,0.02439733346303304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,1024,0.013811555173661975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,768,0.004600000050332812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,768,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,768,0.024754666619830664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,512,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,512,0.012759110993809171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,512,0.02340000040001339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,256,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,256,0.022661333282788593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,256,0.012229333321253458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,128,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,128,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,5120,128,0.020287111401557922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,64,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,5120,32,0.003752888904677497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,64,0.01201600001917945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,5120,32,0.012852444416946836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,65536,0.10809510946273804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,65536,0.20657243993547228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,51200,0.08698578013314141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,65536,0.28219376669989693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,51200,0.1613644493950738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,16384,0.0332275562816196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,16384,0.058541331026289195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,51200,0.22137422031826445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,12288,0.024969778127140466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,16384,0.08615999751620823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,12288,0.04422577884462145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,10240,0.021835555632909138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,12288,0.07072889142566256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,10240,0.037857777542538114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,8192,0.01809955636660258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,10240,0.06237866481145223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,8192,0.03210577699873183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,7168,0.015808888607554965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,8192,0.05342222253481547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,7168,0.029492444462246362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,7168,0.04938488867547777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,6144,0.014147554834683737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,6144,0.027121777335802715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,6144,0.04659466611014473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,5120,0.01239377756913503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,5120,0.024316444993019104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,4096,0.01055555542310079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,5120,0.04301599992646111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,4096,0.02095288866096073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,3584,0.009677333136399588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,4096,0.03831377625465393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,3584,0.01958311100800832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,3584,0.033596442805396184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,3072,0.008967999782827165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,3072,0.018565333551830716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,3072,0.03259822063975864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,2560,0.00833511104186376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,2560,0.01686044368478987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,2560,0.030568000343110826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,2048,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,2048,0.028104000621371802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,2048,0.015071999695565967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,1536,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,1536,0.014174222946166992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,1536,0.026734222968419392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,1024,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,768,0.004622222234805425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,1024,0.013065777719020844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,1024,0.02488800055450863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,512,0.0042239998777707415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,768,0.0244151105483373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,512,0.01272266689274046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,256,0.003584889074166616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,512,0.02269688910908169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,256,0.02267733381854163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,256,0.012068444656001197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,128,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,128,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,64,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,64,0.011714666253990598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,4096,32,0.003522666792074839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,32,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,4096,768,0.012847111456924014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,65536,0.10221778021918403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,4096,128,0.019646222392718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,65536,0.2022657791773478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,51200,0.08083377944098578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,65536,0.2820791138543023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,16384,0.029343111647499934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,51200,0.15761244297027588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,16384,0.0571075545416938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,16384,0.08584977520836724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,12288,0.022245332598686218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,12288,0.04367555512322319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,12288,0.06990310880872938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,10240,0.019339554839664035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,51200,0.2205493317710029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,10240,0.03769866625467936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,8192,0.015554666519165039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,10240,0.06197600232230293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,8192,0.05331733491685656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,8192,0.031522668070263334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,7168,0.014504000544548035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,6144,0.013024888932704926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,7168,0.029474665721257527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,7168,0.04919911093182034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,5120,0.01110577748881446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,6144,0.026949332820044622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,6144,0.04632177617814806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,4096,0.009775111244784461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,5120,0.024235554867320593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,5120,0.04253066579500834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,4096,0.020938666330443487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,4096,0.03770400087038676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,3584,0.019259555472267997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,3584,0.03357955482270982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,3072,0.008389333056079017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,3072,0.01830311119556427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,3072,0.03180177675353156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,2560,0.007501333124107785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,2560,0.01625333395269182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,2560,0.030493332280053034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,2048,0.0069226668112807805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,2048,0.015126221709781222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,2048,0.02786311176088121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,1536,0.005576888720194499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,1536,0.013831111292044321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,1024,0.004767111192146937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,1536,0.027071111732059058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,1024,0.013079110946920184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,768,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,768,0.012727110750145383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,768,0.024367999699380662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,512,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,512,0.01238222254647149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,3584,0.00907466643386417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,512,0.022707555029127333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,256,0.003642666671011183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,256,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,1024,0.024704888463020325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,128,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,256,0.02238133384121789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,128,0.011708444191349877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3584,128,0.019925332731670804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,64,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,64,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,65536,0.093176888095008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3584,32,0.011247110863526663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,65536,0.19964267147911918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,51200,0.07329866621229383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,51200,0.15550933943854436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,65536,0.2827271090613471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3584,32,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,16384,0.030962665875752766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,16384,0.05439022183418274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,12288,0.020807110601001315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,16384,0.0857235524389479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,51200,0.22085955407884386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,12288,0.04264266623391045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,10240,0.017859554953045316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,12288,0.06955199771457247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,8192,0.015080889066060385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,10240,0.036751998795403376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,10240,0.061659554640452065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,7168,0.013311111264758639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,8192,0.031425777408811785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,8192,0.05300177799330818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,6144,0.012072888513406118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,7168,0.02885422110557556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,7168,0.04951111144489712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,5120,0.01072355525361167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,6144,0.026727111803160772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,6144,0.04650222261746725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,4096,0.009545777406957414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,5120,0.023960888385772705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,5120,0.0424213343196445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,3584,0.00868800034125646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,4096,0.020269332660569083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,4096,0.03672711054484049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,3072,0.00795555528667238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,3584,0.018552889426549275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,3584,0.03328888946109348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,2560,0.006948444578382704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,3072,0.03152444296412998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,3072,0.01752444439464145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,2048,0.006232888748248418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,2560,0.016192888220151264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,2560,0.029808001385794744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,1536,0.0053297777970631914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,2048,0.02810844447877672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,1536,0.013820444544156393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,1536,0.026397332549095154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,1024,0.004540444248252445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,1024,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,768,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,768,0.01273155543539259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,1024,0.024863110648261175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,512,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,2048,0.014888889259762235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,768,0.024421332610978022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,256,0.0038071109188927543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,512,0.02234933276971181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,256,0.011963555382357704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,128,0.0035546666218174826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,256,0.022299556268586054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,512,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,128,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,64,0.0034151110384199354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,3072,128,0.01939644416173299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,3072,32,0.0036559998989105225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,64,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,65536,0.07382577657699585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,3072,32,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,51200,0.06026311053170098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,65536,0.19816266165839302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,51200,0.15482044219970703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,16384,0.02677333354949951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,51200,0.220378663804796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,65536,0.2825733290778266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,16384,0.05415911144680447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,12288,0.021948445174429152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,16384,0.0847911106215583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,10240,0.025017778078715008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,12288,0.04186577929390801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,12288,0.06964088810814752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,10240,0.036679112248950534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,8192,0.014331556028789945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,10240,0.06178577740987142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,7168,0.01295377810796102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,8192,0.03115644388728672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,8192,0.053263111246956714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,6144,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,7168,0.02891555428504944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,7168,0.04870488908555773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,5120,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,6144,0.04643199841181437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,6144,0.02685333291689555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,5120,0.023677332533730402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,4096,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,5120,0.042540444268120654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,3584,0.008338666624493068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,4096,0.01978488928741879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,4096,0.03737866547372606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,3072,0.00775999989774492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,3584,0.03441333439615037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,3584,0.018631110588709515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,2560,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,3072,0.03128266665670607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,3072,0.017328000730938382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,2048,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,2560,0.029931555191675823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,2560,0.015978667471143935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,1536,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,2048,0.027789334456125896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,2048,0.014835556348164877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,1536,0.01384888920519087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,1024,0.004509333521127701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,1536,0.02613777750068241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,768,0.004190222256713443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,1024,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,1024,0.024412444896168176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,768,0.012469333079126147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,512,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,768,0.023987554841571387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,256,0.0036311112344264984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,512,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,512,0.02307644486427307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,256,0.011821333732869891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,128,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,256,0.022329777479171753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,64,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,64,0.011109333071443768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2560,32,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2560,128,0.020813332663642038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,32,0.01105955574247572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,65536,0.06223289171854655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,65536,0.1596399943033854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,51200,0.0491920014222463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,65536,0.28181423081292045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,51200,0.12424800131056045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,16384,0.022379555635982092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2560,128,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,51200,0.2211271127065023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,16384,0.04417066772778829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,12288,0.020081778367360432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,16384,0.08501244253582424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,10240,0.01606044504377577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,12288,0.03549422158135308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,12288,0.06991200314627753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,10240,0.030896888838873968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,10240,0.06140622165468004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,8192,0.013615111509958902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,8192,0.02645066711637709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,7168,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,8192,0.05320088730918037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,7168,0.02436977790461646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,6144,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,7168,0.049103111028671265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,6144,0.022465777066018846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,6144,0.04638222191068861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,5120,0.010050666828950247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,5120,0.04016977879736159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,4096,0.00868088917599784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,5120,0.020991111795107525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,4096,0.01740711099571652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,3584,0.00796622203456031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,4096,0.03643466697798835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,3584,0.016192888220151264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,3072,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,3584,0.034027556578318276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,3072,0.015251555376582675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,2560,0.006521777974234686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,3072,0.03193688723776076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,2560,0.014506666196717156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,2048,0.005943111247486538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,2560,0.030250665214326646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,2048,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,1536,0.004941333499219683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,2048,0.02788977821667989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,1536,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,1536,0.026927111877335444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,1024,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,768,0.003979555434650845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,1024,0.02444800072246128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,768,0.0121760004096561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,512,0.00388711111413108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,768,0.02472177810139126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,512,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,256,0.0034151110384199354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,512,0.023031999667485554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,128,0.0032924444725116095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,256,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,256,0.023157333334287006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,64,0.0032755554550223877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,2048,128,0.01958666741847992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,128,0.011651555697123209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,2048,32,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,65536,0.04905422197447883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,64,0.0114702218108707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,2048,32,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,51200,0.040778666734695435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,65536,0.15692267152998182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,65536,0.2814897696177165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,16384,0.01979200045267741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,51200,0.2203751140170627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,51200,0.12212088372972275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,12288,0.016295111841625638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,16384,0.04275911053021749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,16384,0.08381866746478611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,10240,0.014592889282438489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,12288,0.03405155407057868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,12288,0.06879288620418973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,8192,0.012925333446926542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,10240,0.030469334787792627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,10240,0.06176622046364678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,8192,0.02599999970859951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,7168,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,7168,0.02365600069363912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,7168,0.048597332504060536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,6144,0.010739555789364709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,6144,0.04490844408671061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,6144,0.02237333357334137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,5120,0.009460444251696268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,5120,0.040411555104785495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,5120,0.01964000033007728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,4096,0.007870222131411234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,4096,0.035375111632876925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,4096,0.01742044422361586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,3584,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,8192,0.0524888899591234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,3072,0.007114667031500075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,3584,0.0326533317565918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,3584,0.016221332881185744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,2560,0.00647555540005366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,3072,0.03129955463939243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,2560,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,2560,0.029463112354278564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,2048,0.005704888867007361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,2048,0.013577777478430005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,1536,0.004911111046870549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,2048,0.02782755427890354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,1536,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,1536,0.026691555976867676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,1024,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,1024,0.02477688921822442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,3072,0.01517777807182736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,768,0.004259555704063839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,768,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,768,0.02365422248840332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,512,0.0035866668654812705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,512,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,512,0.02207999924818675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,256,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,256,0.011549333731333414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,1024,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,128,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,256,0.02198933396074507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1536,128,0.01958933307064904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,64,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,128,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1536,32,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,64,0.010750222537252637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,65536,0.03682488865322537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1536,32,0.011507555842399597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,65536,0.15509422620137533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,51200,0.031005332867304485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,65536,0.2817395528157552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,51200,0.11967821915944417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,16384,0.014724445011880664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,16384,0.04157155421045091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,51200,0.22141422165764701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,12288,0.013200889031092325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,16384,0.08395822180642022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,10240,0.012688889271683164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,12288,0.03379288978046841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,12288,0.06873955329259236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,8192,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,10240,0.02979822291268243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,10240,0.06129599942101372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,8192,0.02533333334657881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,7168,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,8192,0.0524213347170088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,6144,0.01038577738735411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,7168,0.022960889670583937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,7168,0.04659733176231384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,6144,0.020607110526826646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,5120,0.009001777403884465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,5120,0.018926221463415358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,4096,0.007650666766696506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,5120,0.03978133201599121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,4096,0.016599999533759225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,4096,0.035375111632876925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,3584,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,3584,0.03239555491341485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,3072,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,3584,0.016179554992251925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,6144,0.04353511002328661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,3072,0.014915555715560913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,2560,0.005616888817813661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,3072,0.031727999448776245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,2048,0.005346666607591841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,2560,0.014173333843549093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,2560,0.029790222644805908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,1536,0.004624888714816836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,2048,0.013648000028398303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,2048,0.027781334188249376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,1024,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,1536,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,1536,0.026452443665928308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,1024,0.02368800010946062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,1024,0.012130666938092975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,768,0.003922666526503033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,512,0.003587555554178026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,768,0.02308622168170081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,768,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,256,0.003245333416594399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,512,0.022700443863868713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,512,0.011790222591824003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,128,0.0031635556370019913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,256,0.021958221991856892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,64,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,128,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,1024,32,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,64,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,1024,32,0.01071733319097095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,65536,0.029322667254341975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,65536,0.15427288744184706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,51200,0.028049776951471966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,65536,0.2822924455006917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,51200,0.11875111526913112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,16384,0.013358222113715278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,51200,0.22109333674112955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,1024,128,0.02033866610791948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,16384,0.041478223270840116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,12288,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,12288,0.033379554748535156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,10240,0.010501333408885531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,12288,0.06898844242095947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,10240,0.028740445772806805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,8192,0.009393778112199571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,10240,0.06132266918818156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,8192,0.02405244443151686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,7168,0.00902133352226681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,8192,0.05178577701250712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,16384,0.08415555291705662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,6144,0.010109333528412713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,7168,0.046353777249654136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,6144,0.0207013338804245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,6144,0.04433866673045688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,5120,0.008995555341243744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,5120,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,7168,0.023452444208992854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,4096,0.007334222396214803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,5120,0.04012622104750739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,4096,0.016557332542207506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,3584,0.007076444725195567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,4096,0.03544800149069892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,3584,0.01588000026014116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,3584,0.0329440004295773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,3072,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,3072,0.03080355458789402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,3072,0.015227556228637695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,2560,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,2560,0.029560890462663438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,2560,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,2048,0.013543999857372709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,1536,0.004754666652944353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,2048,0.027898665931489732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,1536,0.012833777401182385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,1024,0.004021333323584663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,2048,0.005066666752099991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,1024,0.012524444195959302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,768,0.003738666574160258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,1024,0.023746665981080797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,768,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,768,0.024040000306235418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,512,0.0035964445107513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,512,0.011525332927703857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,512,0.023123555713229712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,256,0.0033155555526415506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,1536,0.026543999711672466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,128,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,256,0.02197866638501485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,768,128,0.019260444574885897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,64,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,128,0.011510222322411008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,768,32,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,64,0.010827555424637265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,65536,0.024855110380384658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,32,0.011551111108726926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,65536,0.15371999475691053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,51200,0.01945244438118405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,768,256,0.011528000235557556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,65536,0.2814853456285265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,51200,0.11807200643751357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,16384,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,51200,0.22027999824947783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,16384,0.04046666622161865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,12288,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,16384,0.08445422516928779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,12288,0.03163999981350369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,10240,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,12288,0.06789422035217285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,8192,0.009759999811649323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,10240,0.028367112080256145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,10240,0.06079022089640299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,7168,0.009401777552233802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,8192,0.05136177937189738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,7168,0.04674044582578871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,7168,0.022993778189023335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,6144,0.00867377801073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,6144,0.043062223328484424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,5120,0.008049777812427944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,5120,0.018968888454967074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,5120,0.03975288735495673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,8192,0.02440888848569658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,4096,0.007306666837798224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,4096,0.016619554824299283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,4096,0.03498311175240411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,3584,0.007027555671003129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,3584,0.015848888291252982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,3584,0.03258400162061056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,3072,0.006326222171386083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,6144,0.02091111077202691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,3072,0.015152888165579902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,3072,0.030933333767784968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,2560,0.029472000069088403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,2560,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,2048,0.0052986666560173035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,2048,0.013426666458447775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,1536,0.004575110971927643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,1536,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,1536,0.026449778013759192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,1024,0.00397866674595409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,2560,0.005894222193294101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,1024,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,768,0.0035991109907627106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,1024,0.024064888556798298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,768,0.011818666425016193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,2048,0.027454222242037456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,512,0.0034151110384199354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,768,0.024686222275098164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,256,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,512,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,512,0.023432888918452795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,256,0.022014222211307947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,256,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,128,0.011481777661376528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,64,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,512,128,0.01923288901646932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,512,32,0.0029502221279674103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,64,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,65536,0.016001777516471017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,512,32,0.010715555813577441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,65536,0.1528480052947998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,51200,0.01311466677321328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,65536,0.28167910046047634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,51200,0.11726489331987168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,51200,0.22081511550479463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,16384,0.03973955578274197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,16384,0.08286933104197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,12288,0.008434666527642144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,12288,0.03180888957447476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,12288,0.06664533085293241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,10240,0.027802666028340656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,16384,0.009710222482681274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,10240,0.058527999454074435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,8192,0.0240657776594162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,8192,0.04998844530847338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,7168,0.007631111476156447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,10240,0.008719999757077958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,7168,0.021984888447655573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,6144,0.007300444775157505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,7168,0.04561600089073181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,6144,0.020591111646758187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,6144,0.04281422164705065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,5120,0.007948444121413762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,8192,0.00795999997191959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,4096,0.007254222200976477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,5120,0.03943377733230591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,4096,0.01648977730009291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,3584,0.006591111007663939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,4096,0.03501777847607931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,3584,0.015806222955385845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,3072,0.006207110981146495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,3072,0.015136889285511441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,3072,0.030999110804663763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,2560,0.005566222386227713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,5120,0.018975999620225694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,2560,0.02919911013709174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,2048,0.005230222311284807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,3584,0.032227555910746254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,2048,0.013100444442696042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,1536,0.00453599997692638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,2048,0.027427555786238775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,1536,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,1024,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,1536,0.025660443637106154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,1024,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,1024,0.02402666707833608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,768,0.0034977777136696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,2560,0.014344889256689282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,512,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,768,0.023656888140572443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,512,0.011687111523416309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,512,0.02198222279548645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,256,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,256,0.02273688879277971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,128,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,128,0.011364444262451597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,256,128,0.01923111081123352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,64,0.002526222210791376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,64,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,768,0.011821333732869891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,256,32,0.0028159999185138275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,65536,0.014397333065668741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,32,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,65536,0.15039643976423475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,256,256,0.011085333095656501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,51200,0.012883555557992725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,65536,0.28193333413865834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,16384,0.009319999979601966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,51200,0.11554933918846978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,51200,0.22044889132181802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,12288,0.008068444828192392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,16384,0.038120001554489136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,16384,0.08263022369808622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,10240,0.008976889153321585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,12288,0.031176000833511353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,12288,0.06618133518430921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,8192,0.0079644446571668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,10240,0.027800000376171533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,10240,0.05821955866283841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,7168,0.007623111208279927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,8192,0.023665777511066858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,8192,0.049919111861122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,6144,0.007161777880456712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,7168,0.021984888447655573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,7168,0.045560889773898654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,5120,0.007658667034573025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,6144,0.02032977839310964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,6144,0.04279288980695936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,4096,0.0069431112044387394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,5120,0.018580444984965853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,5120,0.03936533464325799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,3584,0.0069271110826068455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,4096,0.03502755694919162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,3584,0.03197333216667175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,3584,0.015594666202863058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,3072,0.006623111251327727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,3072,0.014793778459231058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,4096,0.016583111551072862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,3072,0.030555556217829388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,2560,0.006179555422729916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,2048,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,2560,0.02921066681543986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,2048,0.013073777986897362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,1536,0.004378666480382283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,2048,0.027474666635195415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,2560,0.014557333456145393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,1536,0.01239999963177575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,1536,0.025779555241266888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,1024,0.004104888687531154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,768,0.0037564444873068067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,1024,0.011881777809725868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,1024,0.024096889628304377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,512,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,768,0.023779556155204773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,512,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,256,0.003030222323205736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,512,0.02164711058139801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,768,0.012293332980738746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,128,0.00300444434914324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,256,0.02199999988079071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,256,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,64,0.002817777709828483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,256,128,128,0.018960000740157235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,128,32,0.00249422217408816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,64,0.01072622256146537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,65536,0.01346666696998808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,128,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,128,32,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,51200,0.012180444267061023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,16384,0.009400000174840292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,65536,0.1517982218000624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,12288,0.008292444050312042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,51200,0.11574133237202962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,10240,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,12288,0.030943999687830608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,16384,0.03905511233541701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,8192,0.02414400047726101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,7168,0.006748444504208035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,7168,0.022477333744366963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,6144,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,10240,0.02793244520823161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,5120,0.005997333261701796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,6144,0.0206657780541314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,5120,0.018917333748605516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,4096,0.006620444357395172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,3584,0.006654222392373615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,4096,0.016532444291644625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,3072,0.005995555470387141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,3584,0.01591466698381636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,3072,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,2560,0.006362666686375936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,2048,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,2560,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,8192,0.0069155556460221606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,1536,0.00434400017062823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,1024,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,2048,0.013535111314720578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,1536,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,768,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,1024,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,512,0.0032604444358083936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,256,0.0029279999434947968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,768,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,128,0.0028026666906144884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,512,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,64,0.002522666628162066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,256,0.01145155562294854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,128,0.010760889285140567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,64,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,64,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,65536,0.013171555267439948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,51200,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,64,32,0.010475555227862464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,16384,0.008700444466537898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,65536,0.1503617763519287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,12288,0.008319999608728621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,16384,0.03790311018625895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,51200,0.11552977561950684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,10240,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,8192,0.007116444408893585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,12288,0.031110223796632554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,7168,0.006609777609507243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,8192,0.023754666248957317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,6144,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,10240,0.028239111105600994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,5120,0.006330666856633292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,6144,0.02017688916789161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,4096,0.006593777901596493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,5120,0.01902755598227183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,4096,0.016202666693263583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,3584,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,3072,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,3584,0.015384889311260648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,3072,0.015190222197108798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,2560,0.006502222269773483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,7168,0.021998221675554912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,2048,0.005642666584915585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,2560,0.014467555615637036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,1536,0.004580444345871608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,2048,0.01314399970902337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,1024,0.003958222352796131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,1024,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,1536,0.013022222452693515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,768,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,512,0.003407111184464561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,512,0.011477332976129321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,256,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,256,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,128,0.01073333372672399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,64,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,64,0.010461333725187512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,256,32,32,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,768,0.012208888928095499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,256,32,32,0.010501333408885531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,51200,1.0729733573065865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,51200,1.1883049011230469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,16384,0.39774134423997665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,12288,0.2975262271033393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,16384,0.3431999948289659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,10240,0.24814934200710723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,12288,0.2584284411536323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,10240,0.21600533856285942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,8192,0.2036764489279853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,7168,0.17449066374037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,8192,0.17480888631608751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,6144,0.15308977497948542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,7168,0.15555466545952693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,5120,0.1279511054356893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,6144,0.1360951132244534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,5120,0.11395200093587239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,4096,0.1017155581050449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,3584,0.08969422181447347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,4096,0.09440977705849542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,3072,0.07758844561047025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,3584,0.08439822329415216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,2560,0.06619022289911906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,3072,0.0742480026351081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,2048,0.054824888706207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,2560,0.06529066960016887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,2048,0.05552444524235196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,1536,0.03892088929812113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,1024,0.02787999974356757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,768,0.02276533345381419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,1536,0.04584088921546936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,512,0.01665155589580536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,1024,0.03752977649370829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,256,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,768,0.03180177675353156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,128,0.010859555668301053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,512,0.028307557106018066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,64,0.010431999961535135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,256,0.02498577700720893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,128,0.023397333092159692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,65536,32,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,64,0.023443554838498432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,65536,32,0.022359111242824133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,51200,0.9910977681477865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,65536,1.0811306635538738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,65536,1.2625964482625325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,16384,0.33495023515489364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,12288,0.2479217847188314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,16384,0.27589866850111217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,10240,0.19022044870588514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,12288,0.20885954962836373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,8192,0.1580924457973904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,10240,0.17513510915968153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,7168,0.13393867015838623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,8192,0.1420079999499851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,6144,0.11901244852277969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,7168,0.12674489286210802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,51200,0.8518400192260742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,5120,0.09934310780631171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,6144,0.1109573311275906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,4096,0.08088355594211154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,3584,0.07171289126078288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,5120,0.09326933489905463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,4096,0.07780177725685967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,3072,0.0622346666124132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,2560,0.05335644549793667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,3584,0.06997511121961805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,3072,0.06200622187720405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,2048,0.044343110587861806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,2560,0.054485334290398486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,1536,0.03394755721092224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,2048,0.04697688751750522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,1024,0.025026667449209426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,768,0.020276443825827706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,1536,0.03939822316169739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,1024,0.03203377789921231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,512,0.015058666467666626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,256,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,128,0.009007111191749573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,768,0.02685333291689555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,512,0.023992889457278784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,64,0.007306666837798224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,256,0.021907554732428655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,51200,32,0.007589333587222629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,64,0.020657777786254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,32,0.024403555525673762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,65536,0.41116976737976074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,65536,0.4183031188117133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,51200,0.3237955570220947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,16384,0.11032888624403213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,51200,128,0.0206995556751887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,51200,0.33279111650254994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,12288,0.0826444427172343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,16384,0.11195200019412571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,10240,0.07058311171001859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,8192,0.05823733409245809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,10240,0.0725475549697876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,7168,0.050438221957948476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,8192,0.060623109340667725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,7168,0.054492443799972534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,6144,0.0500311123000251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,5120,0.037852442926830716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,5120,0.04158488909403483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,12288,0.08637955453660752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,4096,0.03149155444569058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,3584,0.028262221150928076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,4096,0.035432888401879206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,3072,0.025006221400366888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,3584,0.03269866771168179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,2560,0.021358221769332886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,6144,0.04455288913514879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,3072,0.02880177895228068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,2048,0.01773422294192844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,1536,0.014104889498816596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,2560,0.025989333788553875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,1024,0.010874666273593903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,2048,0.023031999667485554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,768,0.009454222189055549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,1024,0.018265777164035372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,1536,0.021013332737816706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,512,0.007357333269384172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,768,0.01697422232892778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,256,0.0052897776994440295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,128,0.004567999806669024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,512,0.015957333975368075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,128,0.01421866648726993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,64,0.00461777796347936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,16384,32,0.006492444624503453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,64,0.01444355481200748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,32,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,65536,0.2649262216356066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,51200,0.207458668284946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,65536,0.358242670694987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,16384,256,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,16384,0.07583110862308078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,12288,0.05556622478697035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,51200,0.281514671113756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,16384,0.09768533044391209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,10240,0.04755022128423055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,8192,0.039934221241209246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,12288,0.07645155323876275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,10240,0.06431111362245348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,7168,0.034556445148256086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,8192,0.053542223241594106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,6144,0.03089244498146905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,5120,0.02677511175473531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,7168,0.049850665860705905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,4096,0.022480888499153986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,5120,0.03685599896642897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,4096,0.031025777260462444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,3584,0.01962577799956004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,3072,0.017108443710539076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,3584,0.028449777099821303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,2560,0.015127110812399121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,3072,0.02616888946957058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,2048,0.012934221989578672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,2560,0.024265777733590867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,1536,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,2048,0.021606221795082092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,1536,0.019295111298561096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,6144,0.044585777653588184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,1024,0.008866666919655269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,768,0.007653333246707916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,512,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,1024,0.016600888636377122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,256,0.004949333353175057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,768,0.015615999698638916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,512,0.014819555812411837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,128,0.0041857775714662345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,64,0.003956444561481476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,256,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,12288,32,0.0042364444169733255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,128,0.01348000019788742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,64,0.013539555172125498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,12288,32,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,65536,0.22971733411153158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,51200,0.18015733030107287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,16384,0.06579911046557956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,65536,0.3169982168409559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,51200,0.24912712309095594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,12288,0.048307554589377515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,16384,0.08710222111807929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,10240,0.041482667128245033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,12288,0.0685724417368571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,8192,0.035063998566733465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,10240,0.05773155556784736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,7168,0.030641777647866145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,8192,0.04874577787187365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,6144,0.027042667071024578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,7168,0.04508622156249153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,5120,0.023365333676338196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,6144,0.03973777757750617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,4096,0.018963555494944256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,5120,0.03252977795071072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,3584,0.016920000314712524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,4096,0.028516444895002577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,3072,0.014924445086055331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,3584,0.026313778426912095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,2560,0.013468444347381592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,3072,0.024294222394625347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,2048,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,2560,0.02236266599761115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,1536,0.009798222117953831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,2048,0.020083554916911654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,1024,0.00797333319981893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,1536,0.018218666315078735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,768,0.006956444846259222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,1024,0.015498666299713982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,512,0.00556355549229516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,768,0.01479111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,256,0.004581333448489507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,512,0.014411555396185981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,128,0.0041751112374994485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,256,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,64,0.003921777837806278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,128,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,10240,32,0.004202666795916027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,64,0.012823999755912356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,10240,32,0.01290933373901579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,65536,0.1995288928349813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,51200,0.15713600317637125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,65536,0.27152978049384224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,16384,0.06307555569542779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,12288,0.048435555564032666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,51200,0.21415377987755668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,10240,0.041350222296184964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,16384,0.07639377646976046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,8192,0.03463110989994473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,12288,0.060476442178090416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,7168,0.03014755580160353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,8192,0.0432568887869517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,6144,0.027015111512607996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,7168,0.03901244534386529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,5120,0.020032000210550096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,6144,0.034521778424580894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,4096,0.01717066599263085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,5120,0.028889778587553237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,3584,0.015091554986106025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,4096,0.02557777696185642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,3072,0.013445333474212222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,3584,0.023984000086784363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,2560,0.012411555482281579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,3072,0.02197955548763275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,2560,0.020607110526826646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,2048,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,10240,0.05121155579884847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,1536,0.008692444198661381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,1024,0.0070879997478591064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,2048,0.01867466668287913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,768,0.005551110953092575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,1536,0.016520889268981088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,512,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,1024,0.01479644411140018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,256,0.004530666602982415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,768,0.014422222971916199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,512,0.013462222284740873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,256,0.0127697777416971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,64,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,64,0.012164444559150271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,32,0.003922666526503033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,128,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,8192,32,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,65536,0.1778551075193617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,51200,0.13935022883945042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,65536,0.23640889591640898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,16384,0.050254222419526845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,8192,128,0.0038159998754660287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,51200,0.18780710962083605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,12288,0.037831111086739436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,10240,0.03278133273124695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,16384,0.0675191084543864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,8192,0.02751111156410641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,12288,0.05366577704747518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,7168,0.025777777036031086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,10240,0.045127110348807446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,6144,0.021337777376174927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,8192,0.03770222266515096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,5120,0.019527999891175162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,7168,0.03439911206563314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,6144,0.030903998348448012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,4096,0.016191111670600045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,5120,0.025820444027582806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,3584,0.014966222974989148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,4096,0.02321333355373806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,3072,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,3584,0.02176355487770504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,2560,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,3072,0.02060622142420875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,2048,0.009760888914267222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,2560,0.019130667050679524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,1536,0.008314666648705801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,2048,0.017248888810475666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,1024,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,1536,0.01518311103185018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,768,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,1024,0.014094221923086377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,512,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,768,0.013770666387346057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,256,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,512,0.013440888788965015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,128,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,256,0.012461333639091916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,64,0.0034977777136696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,128,0.012433778080675336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,7168,32,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,64,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,7168,32,0.01312888910373052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,65536,0.17490932676527235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,51200,0.138230217827691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,16384,0.04866577850447761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,65536,0.23302043808831108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,12288,0.036928001377317644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,51200,0.18345333470238578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,10240,0.03221333358022902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,16384,0.06607555680804782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,12288,0.05240799983342489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,8192,0.02685600022474925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,7168,0.02331377731429206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,10240,0.04394222299257914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,8192,0.036566221051745944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,6144,0.0207368897067176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,5120,0.0177528891298506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,7168,0.032891554964913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,4096,0.014446222119861178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,6144,0.03018488817744785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,3584,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,5120,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,4096,0.022832888695928786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,3072,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,2560,0.01076533314254549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,3584,0.021780444516075984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,2048,0.009011555876996782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,3072,0.020371556282043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,1536,0.007703999678293864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,2560,0.018904889623324077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,2048,0.01664355562792884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,1024,0.006254222244024277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,768,0.00526311124364535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,1536,0.01516533394654592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,512,0.004514666481150521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,1024,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,256,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,512,0.013215111361609565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,128,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,768,0.013839999834696451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,64,0.0035128887328836652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,128,0.012112889024946423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,6144,32,0.0034595556143257353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,64,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,32,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,65536,0.1348088847266303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,51200,0.10456889205508763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,65536,0.22801245583428276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,16384,0.03822311096721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,51200,0.18040888839297822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,6144,256,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,12288,0.02882222334543864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,16384,0.06503378020392524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,10240,0.025065778030289546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,12288,0.05048177639643351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,8192,0.021013332737816706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,10240,0.04178399840990702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,7168,0.018367111682891846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,6144,0.016214221715927124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,8192,0.03574311070972019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,5120,0.01419911119672987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,7168,0.032440887557135686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,4096,0.012094222009181976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,6144,0.029884444342719182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,5120,0.025516443782382544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,3584,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,3072,0.010111110905806223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,4096,0.02271733350223965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,2560,0.009387555221716562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,3584,0.02152266601721446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,2048,0.008387555678685507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,2560,0.01755822201569875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,1536,0.006940444310506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,2048,0.01609599921438429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,3072,0.020322667227851022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,1536,0.0148053334818946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,1024,0.005631111148330901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,768,0.00508088908261723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,1024,0.013768889009952545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,512,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,768,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,256,0.003934222377008862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,512,0.013011555704805585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,128,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,256,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,128,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,64,0.003617777592606015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,5120,32,0.004018666843573253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,64,0.011775111158688864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,65536,0.10988266600502862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,5120,32,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,51200,0.09609244267145793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,65536,0.18983022371927896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,16384,0.03657599952485826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,51200,0.15108799934387207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,12288,0.028760890165964764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,12288,0.042876445584827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,10240,0.024736000431908503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,10240,0.03607199920548333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,8192,0.02047288914521535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,16384,0.05454044540723165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,8192,0.031369778845045306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,7168,0.017875555488798354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,6144,0.015858666764365304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,7168,0.02878755662176344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,5120,0.013932444983058505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,5120,0.023058666123284235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,4096,0.012629333469602795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,4096,0.020424000091022916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,3584,0.011531555818186866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,3584,0.019292443990707397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,6144,0.02619733413060506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,3072,0.010671111444632212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,2560,0.009744889206356471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,2048,0.008021333151393467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,3072,0.0174346665541331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,1536,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,2560,0.016015999846988253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,2048,0.014863111906581454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,1024,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,768,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,1536,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,512,0.004600888739029567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,1024,0.013417777915795645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,256,0.004005333201752769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,768,0.01328888866636488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,512,0.013064888616402945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,128,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,64,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,256,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,128,0.011477332976129321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,4096,32,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,64,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,65536,0.10085244311226739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,4096,32,0.012073777616024017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,51200,0.07925511068767972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,16384,0.03228800164328681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,65536,0.18710488743252227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,12288,0.026688888669013977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,51200,0.1485848824183146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,10240,0.02346755564212799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,12288,0.04161244299676683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,16384,0.053583999474843345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,10240,0.035284443034066096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,7168,0.018560000591807894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,8192,0.03073422114054362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,7168,0.02818044357829624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,6144,0.01493866741657257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,5120,0.013827555709415011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,6144,0.02604533400800493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,4096,0.01182577759027481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,5120,0.02269511090384589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,3584,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,4096,0.019820445113711886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,3072,0.00983199973901113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,3584,0.01919644408755832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,2560,0.00872444444232517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,3072,0.01695644524362352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,2048,0.007709333466158972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,2560,0.0162426663769616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,1536,0.0064080001579390625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,2048,0.014898666077189975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,1024,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,1536,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,1024,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,768,0.005050666630268097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,512,0.004229333251714706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,768,0.013136000268989138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,256,0.003588444242874781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,512,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,128,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,256,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,64,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,8192,0.018960888187090557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,128,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3584,32,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,64,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,65536,0.09830488761266072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3584,32,0.011886222494973077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,65536,0.1847564379374186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,51200,0.1459831131829156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,16384,0.03270488977432251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,16384,0.05174755718972948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,12288,0.02068622244728936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,12288,0.04076977901988559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,51200,0.07902044720119901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,10240,0.017471111483044095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,8192,0.015146666102939181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,10240,0.03494666682349311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,7168,0.013693333499961428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,7168,0.028183112541834515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,6144,0.012410666379663678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,6144,0.026161778304311965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,5120,0.022455111145973206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,8192,0.030269334713617962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,4096,0.009382222261693742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,3584,0.00904799997806549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,4096,0.019371555911170114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,3072,0.007992000215583378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,3584,0.018611555298169453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,2560,0.007657777931955125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,3072,0.01664177742269304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,2048,0.00637155564294921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,2560,0.016167110866970487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,5120,0.011128889189826118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,1536,0.005321777943107817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,2048,0.014550222290886773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,1024,0.004552000098758274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,768,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,1024,0.013163555827405719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,512,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,768,0.013074666261672974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,256,0.003306666596068276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,512,0.012499555945396423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,128,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,256,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,128,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,64,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,32,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,1536,0.014456000592973499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,3072,32,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,65536,0.07256266805860731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,51200,0.058424002594418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,65536,0.1834906604554918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,16384,0.025791999366548326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,51200,0.14474933677249485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,12288,0.02068711154990726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,3072,64,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,16384,0.05098311106363932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,10240,0.02122933334774441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,12288,0.040135112073686384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,8192,0.01423111061255137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,7168,0.012724444270133972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,10240,0.03418755531311035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,8192,0.030159999926884968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,6144,0.01166044423977534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,5120,0.010051555931568146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,7168,0.028204444381925795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,6144,0.02555999987655216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,4096,0.009353777600659264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,5120,0.020829333199395072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,3584,0.008563555777072906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,4096,0.018615111708641052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,3072,0.008072000410821702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,3584,0.01760444376203749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,2560,0.007148444652557373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,2048,0.006640888750553131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,3072,0.01665155589580536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,1536,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,2560,0.015527110960748462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,1024,0.004889777965015835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,2048,0.014852444330851236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,1024,0.013034666577974955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,768,0.004264889078007804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,512,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,768,0.012851555314328937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,256,0.00351555562681622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,512,0.012798222402731577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,128,0.0033244445092148254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,256,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,64,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,128,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2560,32,0.003391111062632667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,64,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,65536,0.0653360022438897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,32,0.011933333343929715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2560,1536,0.014160888062583076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,51200,0.053492443429099194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,16384,0.023888889286253188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,65536,0.146053327454461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,12288,0.01959022217326694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,51200,0.11460267172919379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,10240,0.01721333298418257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,16384,0.04078222314516703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,12288,0.0329715543323093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,8192,0.01351999988158544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,10240,0.02847644355561998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,7168,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,8192,0.025080889463424683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,6144,0.011196444431940714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,7168,0.023690667417314317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,5120,0.009726222190592024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,6144,0.021480000681347314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,4096,0.008731555607583787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,5120,0.017955554856194388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,3584,0.007995555798212687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,4096,0.016545777519543965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,3072,0.007646222081449296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,3584,0.015895111693276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,2560,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,3072,0.014849777022997538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,2048,0.005630222045713001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,2560,0.014490667316648694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,1536,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,2048,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,1024,0.004260444392760594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,1536,0.01310488912794325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,768,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,1024,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,512,0.003625777860482534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,768,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,256,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,512,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,256,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,128,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,64,0.0032764443506797156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,32,0.0032426667296224167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,32,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,2048,64,0.01124622258875105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,65536,0.04795022143257988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,51200,0.040213333235846624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,65536,0.14208711518181696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,16384,0.018579555882347953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,51200,0.11232711209191217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,12288,0.016168889072206285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,2048,128,0.0034755555291970572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,10240,0.015204444527626038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,16384,0.0395413339138031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,8192,0.012755555411179861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,12288,0.03259377678235372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,7168,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,10240,0.028212444649802312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,6144,0.011010666688283285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,8192,0.024655999408827886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,5120,0.009758222434255812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,7168,0.02240177823437585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,4096,0.008309333688682979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,6144,0.020610666937298246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,3584,0.0075333333677715724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,5120,0.017900443739361234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,3072,0.007260444263617198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,4096,0.016328889462682936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,2560,0.006594666590293248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,3584,0.015600888265503777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,2048,0.006310222049554189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,3072,0.01484266585773892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,2560,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,1536,0.004925333377387789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,1024,0.004567999806669024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,2048,0.013195555243227216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,768,0.004262222184075249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,1536,0.01297688898113039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,512,0.003584889074166616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,1024,0.012192000117566852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,256,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,768,0.01218488895230823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,128,0.0030008889734745026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,512,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,64,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,256,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1536,32,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,128,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,65536,0.039981332090165876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,64,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1536,32,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,65536,0.14073510964711508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,51200,0.1097795565923055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,16384,0.016568000117937725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,12288,0.012859555582205454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,16384,0.03872355487611558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,10240,0.01178311142656538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,12288,0.03159022331237793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,8192,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,10240,0.02716000046994951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,7168,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,8192,0.023541332946883306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,6144,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,51200,0.03277777632077535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,7168,0.021349334054523047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,5120,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,4096,0.007368889119890001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,5120,0.017550221747822232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,3584,0.0069360000391801195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,4096,0.016210666961140104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,3584,0.015574221809705099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,3072,0.0063235556913746735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,3072,0.014873777826627096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,2560,0.005929777605666055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,2560,0.013869333598348828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,2048,0.005250666704442766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,6144,0.019996444384257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,1536,0.00490488898422983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,2048,0.01314755529165268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,1024,0.004281777888536453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,1536,0.012828444441159567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,768,0.003930666794379552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,1024,0.012178666889667511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,512,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,768,0.012144889268610211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,256,0.0032631111227803757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,512,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,128,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,256,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,64,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,128,0.011139555937714048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,1024,32,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,64,0.01110577748881446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,65536,0.030206223328908283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,1024,32,0.010825778047243753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,51200,0.028000000450346205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,65536,0.13914755980173746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,16384,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,51200,0.10877688725789388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,12288,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,16384,0.0380311111609141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,10240,0.010627555350462595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,12288,0.030921777089436848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,8192,0.009720889230569204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,10240,0.026178666286998328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,7168,0.00925244473748737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,8192,0.023013333479563396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,6144,0.008369777765538957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,7168,0.021350221501456365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,5120,0.008761777646011775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,6144,0.01994400057527754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,4096,0.007631111476156447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,5120,0.01756977703836229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,3584,0.0069288888739215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,4096,0.016184889607959323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,3072,0.006412444429265128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,3584,0.015526221858130561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,2560,0.005650666852792104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,3072,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,2048,0.0052577778697013855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,2560,0.01385155568520228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,1536,0.004582222137186262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,2048,0.01313244468635983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,1024,0.004107555581463708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,1536,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,768,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,1024,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,512,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,768,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,256,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,512,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,128,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,256,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,64,0.0029786665820413162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,128,0.011134222149848938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,768,32,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,64,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,65536,0.02418488926357693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,768,32,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,51200,0.020181333025296528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,65536,0.13824710581037733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,16384,0.012850667039553324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,51200,0.10810577869415283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,12288,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,16384,0.0377146667904324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,10240,0.009945777555306753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,12288,0.030351112286249798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,8192,0.010024000373151567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,10240,0.026087111896938745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,7168,0.009465778039561378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,8192,0.022992889086405437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,6144,0.008435555630260045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,7168,0.021359999974568684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,5120,0.008007110820876228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,6144,0.020000888241661918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,4096,0.0074933336840735535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,5120,0.01754844519827101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,3584,0.0069422221018208405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,4096,0.016174222032229107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,3072,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,3584,0.015491555134455362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,2560,0.005903999838564131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,3072,0.014759999182489185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,2048,0.005323555734422471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,2560,0.013836444252067141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,1536,0.004936000125275718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,2048,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,1024,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,1536,0.012741333080662621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,768,0.003966222206751506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,1024,0.012110222544935016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,768,0.012164444559150271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,512,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,512,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,256,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,256,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,128,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,128,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,64,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,64,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,512,32,0.0029253332565228143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,512,32,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,65536,0.01553600033124288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,65536,0.13741778002844915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,51200,0.016526222229003906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,51200,0.10690310928556655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,16384,0.009718221922715506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,12288,0.009689778089523315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,12288,0.029634667767418757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,10240,0.009008888569143083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,16384,0.03744355506367154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,8192,0.008285333712895712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,10240,0.02611644400490655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,7168,0.008038221961922115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,8192,0.023026666707462732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,6144,0.007627555893527136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,7168,0.021349334054523047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,5120,0.008336000144481659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,6144,0.01996888882584042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,5120,0.017535110314687092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,4096,0.007664889097213745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,3584,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,4096,0.01614755557643043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,3072,0.00638488887084855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,2560,0.005656888915432825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,3584,0.015758223003811307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,3072,0.014514666464593677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,2048,0.005281777845488654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,1536,0.004552888787455029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,2560,0.014239999983045789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,1024,0.003936888857020272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,1536,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,768,0.003527111063400904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,1024,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,512,0.003264888914095031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,768,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,256,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,512,0.01185066666867998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,128,0.0029377777957253983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,256,0.011844444606039258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,64,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,128,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,256,32,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,64,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,2048,0.01311466677321328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,256,32,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,65536,0.014510222607188754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,51200,0.01460444430510203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,16384,0.009472889204819998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,65536,0.13577244016859266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,12288,0.009415999882751042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,16384,0.036357333262761436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,51200,0.10465688837899102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,10240,0.008719999757077958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,8192,0.00943199959066179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,12288,0.029623111089070637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,7168,0.00905688852071762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,10240,0.026292444931136236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,6144,0.008366222182909647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,8192,0.022648000054889258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,5120,0.008365333080291748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,7168,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,4096,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,6144,0.019944889677895438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,5120,0.01752000053723653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,3584,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,4096,0.01588711142539978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,3072,0.005987555616431766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,2560,0.005676444619894028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,3584,0.015494222442309061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,2048,0.004912888838185204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,3072,0.014466666513019137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,1536,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,2048,0.013088000317414602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,2560,0.013916444447305469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,1024,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,768,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,1024,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,768,0.012310222619109683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,512,0.003165333428316646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,256,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,512,0.011810666984981961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,128,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,256,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,64,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,128,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,128,32,0.0028328889360030494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,64,0.011363555159833698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,32,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,128,1536,0.012608889076444836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,65536,0.02036355601416694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,51200,0.01646844380431705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,65536,0.13604622417026097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,16384,0.009081777599122789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,51200,0.10595911078982884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,16384,0.036017778846952654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,12288,0.009422221945391761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,10240,0.008903111020723978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,12288,0.029633777009116277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,8192,0.008035555481910706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,10240,0.026132444540659588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,7168,0.008952000074916417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,8192,0.022709333234363135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,6144,0.008370666868156856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,7168,0.021446221404605444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,6144,0.019718221492237516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,5120,0.007720888488822513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,4096,0.0069857777820693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,5120,0.017543110582563613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,3584,0.006610666712125142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,4096,0.016210666961140104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,3072,0.006246222390068903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,3584,0.015605333778593274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,2560,0.0052915554907586836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,3072,0.0148062225845125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,2560,0.013836444252067141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,2048,0.004976888911591636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,1536,0.004546666724814309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,2048,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,1024,0.003973333372010125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,1536,0.013170666992664337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,768,0.0035777779089079965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,1024,0.012129777835475074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,512,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,512,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,768,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,256,0.0029226665695508323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,256,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,128,0.011319999893506369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,64,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,64,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,64,0.011171555353535546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,65536,0.04221066501405504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,64,32,0.011380444798204633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,51200,0.03396888905101352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,16384,0.013893332746293811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,65536,0.13472889529334173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,51200,0.10519199901156956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,16384,0.03640088770124648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,12288,0.009481777747472128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,10240,0.008396444221337637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,8192,0.008039111064540016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,10240,0.025888888372315302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,12288,0.02983733349376255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,7168,0.00868888861603207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,8192,0.022706665926509436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,6144,0.008014221986134848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,5120,0.0076657773719893555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,6144,0.019704888264338177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,7168,0.021345777644051447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,4096,0.007304888632562425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,5120,0.01726222203837501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,3584,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,3072,0.005931555396980709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,4096,0.01616533266173469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,3584,0.01513155632548862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,2560,0.005562666803598404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,3072,0.014579556054539151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,2048,0.004958222309748332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,1536,0.00453333349691497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,2560,0.013491555220550962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,1024,0.0038497779104444715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,2048,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,768,0.003626666549179289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,1536,0.013097777962684631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,512,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,1024,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,256,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,768,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,128,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,512,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,256,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,64,0.002757333426011933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,128,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,225,32,32,0.002573333266708586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,64,0.010446222292052375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,225,32,32,0.010760000182522668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,51200,1.1826658248901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,51200,1.0776746537950304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,16384,0.33970043394300675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,16384,0.39528976546393496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,12288,0.29748977555169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,12288,0.2587973276774089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,16384,0.2904968791537815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,10240,0.26191287570529515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,12288,0.22539199723137748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,10240,0.21543199486202666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,8192,0.21476622422536215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,8192,0.1744248867034912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,10240,0.19218044810824922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,51200,0.8549840185377332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,7168,0.18492889404296875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,8192,0.1573022206624349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,7168,0.1557111077838474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,6144,0.16187822818756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,6144,0.1362204419242011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,7168,0.1416977776421441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,5120,0.13545510503980848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,5120,0.11386044820149739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,4096,0.10929066605038112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,6144,0.12596444288889566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,4096,0.09499910804960464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,3584,0.09645955430136786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,3584,0.08403644296858047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,4096,0.09281688928604126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,3072,0.0840266678068373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,3584,0.08440622356202866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,3072,0.07420800129572551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,2560,0.0719475547472636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,5120,0.10954311158921982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,3072,0.07710311147901747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,2560,0.06536444690492418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,2048,0.0595724450217353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,2560,0.06926133235295613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,1536,0.04569688770506117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,2048,0.06193866994645861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,2048,0.05603466431299845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,1536,0.04549955659442478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,1024,0.03269866771168179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,1024,0.03703644540574815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,768,0.026484444737434387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,1024,0.04786577820777893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,768,0.03156355685657925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,512,0.01909955508179135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,768,0.04324355390336779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,512,0.02684088879161411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,256,0.011331555744012197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,512,0.03971644573741489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,256,0.02425066630045573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,256,0.03515555461247762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,128,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,128,0.028975112570656672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,64,0.010443555812040964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,128,0.02369511127471924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,65536,32,0.009349333743254343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,64,0.022729777627521094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,65536,32,0.02347822156217363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,65536,1536,0.05470488799942864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,65536,1.2625608444213867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,65536,1.0745147069295247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,51200,0.991636488172743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,51200,0.8501271141899956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,16384,0.3340604570176866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,16384,0.27522844738430446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,16384,0.2549831072489421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,12288,0.24793778525458443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,12288,0.20969600147671172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,10240,0.20204533471001518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,65536,0.9542604022555881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,10240,0.17509333292643228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,51200,0.7448995378282336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,12288,0.19710311624738905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,8192,0.16664977868398032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,8192,0.14217778046925864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,10240,0.16938311523861357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,7168,0.14241955015394422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,7168,0.12654399871826172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,6144,0.1250071128209432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,6144,0.111117336485121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,8192,0.13878222306569418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,5120,0.10444000032212998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,7168,0.1253875494003296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,5120,0.09347377883063422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,6144,0.11253778139750163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,4096,0.07763377825419109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,5120,0.09708178043365479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,3584,0.07518577575683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,3584,0.06978755527072482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,3072,0.06542844242519803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,4096,0.08337689108318752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,3584,0.07597155703438653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,2560,0.05631733602947659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,3072,0.06203022268083361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,4096,0.08506133159001668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,2560,0.05447466505898369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,2048,0.04640711016125149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,3072,0.07055289215511747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,2560,0.0625964469379849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,1536,0.03644711110326979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,2048,0.0468444459968143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,1536,0.03923644291030036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,2048,0.05649244123035007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,1024,0.02608000073168013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,768,0.020929776959949072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,1536,0.04993777804904514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,1024,0.03135910961363051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,1024,0.04371111260520088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,512,0.014714666538768344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,768,0.026125333375400964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,512,0.0233262214395735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,768,0.04053511222203573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,256,0.009720000127951304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,256,0.0326186650329166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,128,0.009317333499590555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,256,0.022097777989175584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,128,0.0207315550910102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,64,0.007346666521496243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,128,0.026605332891146343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,51200,32,0.007675555845101674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,64,0.020634666085243225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,51200,512,0.03558844327926636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,51200,32,0.02085688875781165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,65536,0.37447378370496964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,51200,0.29502399762471515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,65536,0.4187324311998155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,51200,0.3294862111409505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,16384,0.10430222087436253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,16384,0.11187822288937038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,16384,0.11966489420996772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,65536,0.41268711619906956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,51200,0.32640443907843697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,12288,0.07610755496554904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,10240,0.06496710909737481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,12288,0.08676444159613715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,10240,0.07273599836561415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,12288,0.09591555595397949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,8192,0.05527822176615397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,8192,0.060144000583224826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,7168,0.04692622356944614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,10240,0.08431910806232029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,8192,0.07158133056428698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,6144,0.04159288936191135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,7168,0.05449777841567993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,7168,0.06735999716652764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,6144,0.04986933204862806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,6144,0.061526219050089516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,5120,0.03532888823085361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,4096,0.029430223835839167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,5120,0.04103733433617486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,5120,0.05462400118509928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,3584,0.02628355556064182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,4096,0.048094222942988075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,4096,0.03603466682963901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,3584,0.03235733177926805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,3072,0.022991999983787537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,3584,0.045174221197764076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,2560,0.020082667469978333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,3072,0.028931554820802476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,3072,0.042971554729673594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,2560,0.026176888081762526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,2560,0.039092444711261325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,2048,0.023381332556406658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,2048,0.03725511166784499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,1536,0.013395555317401886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,1536,0.020617778102556866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,1536,0.03446933295991685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,1024,0.01073688848151101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,1024,0.018366222580273945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,1024,0.03124622172779507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,768,0.009146666361225976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,768,0.016582222448454965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,768,0.029194666279686823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,512,0.007477333148320516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,2048,0.016680000556839835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,256,0.0053102220926019884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,512,0.026033777329656813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,512,0.01556622154182858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,128,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,256,0.024432889289326135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,128,0.014170666535695394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,256,0.014836443795098199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,64,0.004202666795916027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,16384,128,0.021337777376174927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,16384,32,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,32,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,16384,64,0.014138667119873894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,65536,0.2858160071902805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,51200,0.22499465942382812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,65536,0.35590044657389325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,51200,0.2807226710849338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,16384,0.07982755369610257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,16384,0.09755555788675944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,65536,0.41095911131964785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,51200,0.3232453399234348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,16384,0.11903377374013265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,12288,0.05956800116433037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,10240,0.05062399970160591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,12288,0.07626222239600287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,12288,0.09570844305886163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,8192,0.04256177941958109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,10240,0.06459466616312663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,10240,0.08362133635414971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,7168,0.036730666955312095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,8192,0.053746667173173696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,8192,0.07167644633187188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,7168,0.0496782230006324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,6144,0.03323466579119364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,7168,0.06617599725723267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,5120,0.028500444359249536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,6144,0.04453688859939575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,6144,0.06158577733569675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,5120,0.036609778801600136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,4096,0.02397955622937944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,5120,0.05419466561741299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,4096,0.03107644451989068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,3584,0.021180444293551978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,4096,0.0484168893761105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,3584,0.028204444381925795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,3072,0.018499554859267343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,3584,0.04557866520351834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,3072,0.02648711038960351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,2560,0.01592088904645708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,3072,0.04267288910018074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,2560,0.023787554767396715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,2048,0.013717333475748697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,2560,0.039763556586371526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,1536,0.011523555550310345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,2048,0.02141688929663764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,2048,0.03680266605483161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,1024,0.009000889129108852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,1536,0.03375466664632162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,1536,0.019533332851197984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,768,0.008164444731341468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,1024,0.03031733300950792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,1024,0.01705422169632382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,512,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,768,0.02714666724205017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,768,0.01587022178702884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,256,0.004944889081848992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,512,0.01441422270403968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,512,0.025819554924964905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,128,0.00425511101881663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,256,0.02349244389269087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,256,0.013847110999955071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,128,0.013134222063753339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,64,0.0042364444169733255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,12288,32,0.004503111044565837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,12288,128,0.021536888347731695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,64,0.01311377767059538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,12288,32,0.013767999907334646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,65536,0.22922221819559732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,65536,0.31625599331325954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,51200,0.17835644880930582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,51200,0.2492471006181505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,16384,0.06526488727993436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,16384,0.08723022540410359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,65536,0.41069510247972274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,51200,0.32100534439086914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,16384,0.11791377597384983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,12288,0.04830044507980347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,10240,0.041456000672446355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,12288,0.06840088632371691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,12288,0.09467111031214397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,8192,0.034712887472576566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,10240,0.057966219054328076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,10240,0.08305155568652682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,7168,0.03049955434269375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,8192,0.04853333367241753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,8192,0.07113689184188843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,7168,0.04482400086190966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,6144,0.02699822187423706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,7168,0.0657679968410068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,6144,0.03930755456288656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,5120,0.023919110496838886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,6144,0.0607831080754598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,5120,0.03259911139806112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,4096,0.019184889064894784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,5120,0.05316533313857185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,3584,0.017148445049921673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,4096,0.02830666634771559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,3584,0.026126222478018865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,4096,0.047503110435273915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,3072,0.015322667029168872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,3584,0.04389333393838671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,2560,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,3072,0.041364444626702204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,3072,0.024841777152485315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,2560,0.022026666336589392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,2048,0.01183111137813992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,2560,0.038118223349253334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,1536,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,2048,0.01995466649532318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,2048,0.036679999695883855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,1536,0.017973333597183228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,1024,0.008531555533409119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,1536,0.03293422195646498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,768,0.006672888994216919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,1024,0.015781333049138386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,1024,0.02992266747686598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,768,0.014840889308187695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,512,0.0058053334554036455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,768,0.027839110957251653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,256,0.004811555561092165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,512,0.013743110828929476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,256,0.013146667016877068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,512,0.02681333323319753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,128,0.004276444514592488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,256,0.023336889015303716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,64,0.0038328886859946777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,10240,128,0.020968000094095867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,128,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,10240,32,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,64,0.012768000364303589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,65536,0.19128710693783232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,10240,32,0.013418667018413544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,65536,0.27156088087293834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,51200,0.1506062216228909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,51200,0.2140960031085544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,65536,0.28195201026068795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,16384,0.05477777785725064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,16384,0.07682933410008748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,51200,0.22281511624654135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,12288,0.04087111022737291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,16384,0.08926133314768474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,12288,0.0606879989306132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,10240,0.03462933169470893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,12288,0.07324444585376315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,10240,0.05116533239682516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,8192,0.02956533432006836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,10240,0.06394755840301514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,8192,0.04327999883227878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,7168,0.025570667452282373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,8192,0.05494933327039083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,7168,0.03886222177081638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,6144,0.022675555613305833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,7168,0.05154933200942146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,6144,0.03418488966094123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,5120,0.019723556107944913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,6144,0.047805332475238375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,4096,0.016167999969588388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,5120,0.04239022069507175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,5120,0.029669334491093952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,3584,0.01441511180665758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,4096,0.03800977932082282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,4096,0.026103999879625108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,3072,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,3584,0.036172442966037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,3584,0.024296889702479046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,2560,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,3072,0.03469511204295688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,3072,0.0225564440091451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,2560,0.020642666353119742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,2560,0.03243644369973077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,2048,0.019048889478047688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,2048,0.031473779016070895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,1536,0.008640000389681922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,1536,0.016223111086421542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,1024,0.007292444507280986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,1536,0.02826489011446635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,1024,0.014876445134480795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,1024,0.02555466691652934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,768,0.0057955553962124725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,768,0.024880889389250014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,768,0.014158222410413953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,512,0.004303110970391167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,2048,0.010244444840484196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,256,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,512,0.013515555196338229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,256,0.023360000716315374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,256,0.013125333521101209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,128,0.0035546666218174826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,128,0.02033511135313246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,64,0.0034062222888072333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,128,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,8192,32,0.0037466668420367767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,64,0.012635555532243518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,8192,32,0.01348355578051673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,65536,0.1756817764706082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,8192,512,0.024068444967269897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,51200,0.13910755846235487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,65536,0.2378622161017524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,51200,0.18667111131880018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,16384,0.05083822210629781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,16384,0.06755733489990234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,65536,0.2770480049981011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,12288,0.038345777326160006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,51200,0.21942578421698677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,16384,0.08735999796125625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,12288,0.05392711030112373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,10240,0.03283111254374186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,10240,0.04553066690762838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,10240,0.06267466810014513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,8192,0.02751911183198293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,8192,0.03791200121243795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,8192,0.05371555685997009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,7168,0.034416000048319496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,7168,0.05055377880732218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,6144,0.02251111136542426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,6144,0.030953778160942927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,6144,0.04704088966051737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,5120,0.01920977731545766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,7168,0.02646666599644555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,5120,0.042167110575569995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,4096,0.016698666744761996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,4096,0.023456000619464453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,4096,0.037490665912628174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,12288,0.0714506639374627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,3584,0.015263110399246216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,3584,0.036093334356943764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,3584,0.022269333402315777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,3072,0.013072888884279462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,5120,0.026446221603287592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,3072,0.020709334148301017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,2560,0.011583111352390714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,3072,0.0343466666009691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,2048,0.00980000032318963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,2560,0.03257866700490316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,2560,0.019282667173279654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,1536,0.008341333104504479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,2048,0.031101332770453557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,1536,0.015593777100245157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,1536,0.027230223019917805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,1024,0.0063182223174307085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,1024,0.026290666725900438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,1024,0.01422044469250573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,768,0.006314666734801398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,768,0.013537777794731988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,512,0.0052693333062860705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,768,0.02510133385658264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,2048,0.017332444588343304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,256,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,512,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,512,0.025052444802390203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,128,0.0041911109454101985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,256,0.012703110774358114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,256,0.02200444373819563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,64,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,7168,128,0.020614221692085266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,128,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,7168,32,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,64,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,65536,0.17628799544440377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,7168,32,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,65536,0.23164267010158965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,51200,0.13788978258768717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,51200,0.18277955055236816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,16384,0.048078222407235034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,65536,0.27359377013312447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,16384,0.06620711088180542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,51200,0.21695378091600206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,12288,0.036978665325376726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,16384,0.0862026678191291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,10240,0.03202133377393087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,12288,0.07043111324310303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,12288,0.052539553907182485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,8192,0.02658844490845998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,10240,0.04377333323160807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,10240,0.062365333239237465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,7168,0.023249778482649062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,8192,0.03717155588997735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,7168,0.03299555513593886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,6144,0.020724443925751578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,7168,0.05033066537645128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,6144,0.03013955553372701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,6144,0.04717155628734165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,5120,0.017848889032999676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,5120,0.04186844494607714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,4096,0.014544889330863953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,4096,0.02331377731429206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,8192,0.05297155512703789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,3584,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,4096,0.03800977932082282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,3584,0.021735999319288466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,3072,0.01164533363448249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,3584,0.035982221364974976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,3072,0.03447733322779337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,2560,0.0103511114915212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,5120,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,2560,0.01885511146651374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,2048,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,2560,0.032454222440719604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,2048,0.016940444707870483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,1536,0.008079110748238033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,2048,0.030403554439544678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,1536,0.014883556299739413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,1024,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,3072,0.020283554991086323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,1536,0.027609777119424608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,1024,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,768,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,1024,0.024855999483002558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,768,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,512,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,768,0.025217778152889673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,512,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,256,0.012379555238617791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,256,0.02275200022591485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,128,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,128,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,128,0.019631110959582858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,64,0.0034711110509104202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,6144,512,0.024078221784697637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,32,0.003177777760558658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,64,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,65536,0.13265866703457302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,65536,0.2273191081153022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,6144,256,0.004023999803596073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,51200,0.11352266205681695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,6144,32,0.012172444827026792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,65536,0.2715982331169976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,51200,0.17605333858066133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,16384,0.03903466794225905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,51200,0.21490844090779623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,16384,0.06537955337100558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,12288,0.029203553994496662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,16384,0.08509333266152276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,12288,0.051144000556733876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,12288,0.06891644663280912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,10240,0.025343111819691126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,8192,0.020989333589871723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,10240,0.06140622165468004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,10240,0.042783111333847046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,7168,0.01939022209909227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,8192,0.03566133313708835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,8192,0.052440888351864286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,6144,0.016961778203646343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,7168,0.04986044433381823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,7168,0.032864888509114586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,5120,0.01510933372709486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,6144,0.029787556992636785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,6144,0.046462221278084644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,4096,0.012552888856993781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,5120,0.04149511125352647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,5120,0.02571288910177019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,4096,0.022944889134830896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,3584,0.012002666791280111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,4096,0.03724088933732774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,3072,0.010237333675225576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,3584,0.02148799929353926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,3584,0.03566133313708835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,3072,0.020000000794728596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,3072,0.03386933273739285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,2560,0.018065777089860704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,2560,0.031888888941870794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,2048,0.00835111074977451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,2048,0.015831111205948722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,2048,0.02966844373279148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,1536,0.0075235557225015424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,1536,0.01482488877243466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,1536,0.02679288884003957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,1024,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,1024,0.01386844449573093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,1024,0.025380444195535447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,768,0.004893333133724001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,2560,0.009743111001120673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,512,0.004607111215591431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,768,0.02439911166826884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,512,0.02368444369898902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,256,0.0038720000949170855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,256,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,256,0.022680888573328655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,128,0.003644444462325838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,128,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,768,0.01311377767059538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,64,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,5120,128,0.020029332902696397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,5120,32,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,512,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,64,0.011843555503421359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,5120,32,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,65536,0.11193688710530598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,51200,0.086444444126553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,65536,0.1916959948009915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,51200,0.15130221843719482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,65536,0.26088089413113064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,16384,0.036080890231662326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,51200,0.20613600148095024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,12288,0.028784887658225164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,16384,0.055029332637786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,16384,0.08203644222683378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,12288,0.04169333312246534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,10240,0.02472977836926778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,12288,0.06656533479690552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,8192,0.020303999384244282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,10240,0.035749332772360906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,10240,0.05945599741405911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,7168,0.017702221870422363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,8192,0.051199999120500356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,7168,0.0487333337465922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,7168,0.028697778781255085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,6144,0.01585155559910668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,6144,0.045500444041358105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,6144,0.026406221919589575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,5120,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,5120,0.04064711266093784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,5120,0.02276177704334259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,4096,0.010182222558392419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,8192,0.030577778816223145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,3584,0.009850666754775578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,4096,0.020655110478401184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,4096,0.03665688965055678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,3072,0.00905688852071762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,3584,0.01917155583699544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,3584,0.03503466645876566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,3072,0.017480888300471835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,3072,0.033349331882264875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,2560,0.0158942225906584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,2560,0.03046400017208523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,2048,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,2048,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,2048,0.028047998746236164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,1536,0.00599911105301645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,1536,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,2560,0.008163555628723567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,1536,0.026504889130592346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,1024,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,1024,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,768,0.0042462220622433555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,1024,0.025798221429189045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,768,0.0127688886390792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,512,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,768,0.02479644450876448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,512,0.012502222425407834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,256,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,512,0.0247697780529658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,256,0.01182577759027481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,128,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,256,0.02328799996111128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,128,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,64,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,4096,128,0.02033866610791948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,4096,32,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,64,0.011807999677128263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,65536,0.10142844253116184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,4096,32,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,65536,0.1872782177395291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,51200,0.07968888680140178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,51200,0.1485439936319987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,65536,0.2608773443433973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,16384,0.032427555984920926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,16384,0.08129333125220405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,16384,0.05335466729270088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,12288,0.02660088903374142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,12288,0.06599466668234931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,12288,0.04126133190260993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,10240,0.02458311120669047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,10240,0.05883644686804878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,10240,0.035381333695517644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,8192,0.019934222102165222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,8192,0.05098755492104424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,8192,0.030617776844236586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,7168,0.016511110795868766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,51200,0.20507821771833631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,7168,0.0483511123392317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,6144,0.015853333804342482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,7168,0.028322666883468628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,5120,0.013614222407341003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,6144,0.026203556193245783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,6144,0.04517155554559496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,4096,0.011914666328165265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,5120,0.02269511090384589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,5120,0.04010755485958523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,3584,0.010747555229398938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,4096,0.020410666863123577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,4096,0.03601333167817857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,3584,0.01891377733813392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,3584,0.03275733192761739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,3072,0.01720622181892395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,2560,0.008795555267069075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,3072,0.03144711256027222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,2560,0.016167999969588388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,2048,0.007679111427730984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,2560,0.029217776325013902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,2048,0.01479200025399526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,1536,0.00646577775478363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,2048,0.028000000450346205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,1536,0.013794666363133324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,1536,0.026872000760502283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,1024,0.005365333209435145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,3072,0.009198221895429824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,768,0.004624888714816836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,1024,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,768,0.012891555825869242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,512,0.004284444368547863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,512,0.012744888663291931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,512,0.024452444579866197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,256,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,1024,0.025087111526065405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,256,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,768,0.024057777391539678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,128,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,128,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,64,0.003024000053604444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,64,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3584,32,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3584,32,0.01185155577129788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,65536,0.09868088695738052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,256,0.02237777743074629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,65536,0.1845759948094686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,51200,0.07872622542911105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3584,128,0.02001333402262794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,65536,0.26038845380147296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,51200,0.14579466978708902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,16384,0.03307466705640157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,51200,0.20505955484178331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,16384,0.05196088883611891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,16384,0.08081777890523274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,12288,0.027119111683633592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,10240,0.020003555549515616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,12288,0.04028355413013034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,12288,0.06605866882536146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,10240,0.05849777989917331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,10240,0.03463733196258545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,8192,0.016034666034910414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,8192,0.03014844324853685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,8192,0.050781332784228854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,7168,0.014881778094503613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,7168,0.02813777658674452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,7168,0.04804977774620056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,6144,0.01311466677321328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,6144,0.02590488890806834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,6144,0.045215109984079994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,5120,0.011508444117175208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,5120,0.022244445151752893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,5120,0.04005777835845947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,4096,0.009561777942710454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,4096,0.01979733341270023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,4096,0.034323556555642024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,3584,0.009385777844323052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,3072,0.008349333372380998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,3584,0.03295733200179206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,3584,0.018207111292415194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,2560,0.007400000260935889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,3072,0.016778666112158034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,3072,0.030993776188956365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,2048,0.006088888893524806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,2560,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,2560,0.0288355549176534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,1536,0.005336888962321811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,2048,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,2048,0.027757333384619817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,1024,0.004546666724814309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,1536,0.013678222066826291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,1536,0.02609777781698439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,768,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,1024,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,1024,0.02405333353413476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,512,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,768,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,768,0.023762666516833838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,256,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,512,0.012161778079138862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,512,0.0233271105421914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,128,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,256,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,256,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,64,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,128,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,3072,32,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,3072,128,0.019908444748984445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,65536,0.07140888770421346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,64,0.011425777441925473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,3072,32,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,51200,0.0583768884340922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,65536,0.1823786629570855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,51200,0.14450044102138943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,16384,0.025257777836587694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,65536,0.26037687725490993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,51200,0.20477867126464844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,16384,0.050035556157430015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,16384,0.08064977990256415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,12288,0.019694222344292533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,10240,0.022981334063741896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,12288,0.06596000327004327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,10240,0.03427199853791131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,8192,0.014102222190962898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,10240,0.05867377916971842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,8192,0.030088000827365454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,7168,0.012727999852763282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,7168,0.027795556518766616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,7168,0.04789066645834181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,6144,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,12288,0.03986577855216132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,6144,0.02570399973127577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,5120,0.010415111151006486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,6144,0.04504533277617561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,5120,0.021335111724005804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,4096,0.010022222167915767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,5120,0.04004266526963975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,8192,0.05045066608322991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,3584,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,4096,0.018629334039158292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,4096,0.033909334076775446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,3072,0.008319999608728621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,3584,0.017887110511461895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,3584,0.033160001039505005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,2560,0.007284444239404466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,3072,0.016549333930015564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,3072,0.03087377879354689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,2048,0.006613333192136552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,2560,0.01550222271018558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,2560,0.029117332564459905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,1536,0.005562666803598404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,2048,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,2048,0.027815110153622095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,1024,0.004826666580306159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,1536,0.013779555757840475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,1024,0.024731554918819006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,1024,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,768,0.004520888957712385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,768,0.012447111308574677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,768,0.02401155564520094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,512,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,512,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,512,0.0229759994480345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,1536,0.025773333178626165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,256,0.0035048888789282907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,256,0.022320000661744013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,128,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,128,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2560,128,0.01925244430700938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,64,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,64,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2560,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,32,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,65536,0.0634595553080241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2560,256,0.011712888876597086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,65536,0.14430133501688638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,51200,0.048972444401847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,65536,0.26060088475545246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,51200,0.1144755548901028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,16384,0.023759111762046814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,51200,0.2047875589794583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,16384,0.04048177931043837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,12288,0.019496000475353666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,16384,0.0802604423628913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,10240,0.017102221647898357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,12288,0.03292355603641934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,12288,0.06562844249937269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,8192,0.013447999954223633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,10240,0.028628445333904688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,10240,0.0586106644736396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,7168,0.012167111039161682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,8192,0.025411556164423626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,7168,0.023360000716315374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,8192,0.05075822273890177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,6144,0.011031111081441244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,7168,0.0477751096089681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,5120,0.009860444400045607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,6144,0.04478844337993198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,6144,0.02239199976126353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,5120,0.018579555882347953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,5120,0.03874844312667847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,4096,0.016846223009957206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,3584,0.008016889293988546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,4096,0.034450666771994695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,3584,0.015796444482273526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,3072,0.0073626670572492815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,3072,0.01512977812025282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,3072,0.03068888849682278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,2560,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,4096,0.009085333181752099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,2560,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,2048,0.006086222413513396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,2560,0.029193778832753498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,2048,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,3584,0.03232177760865953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,1536,0.004957333207130432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,2048,0.028833776712417603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,1024,0.004269333349333869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,1536,0.012793777717484368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,1024,0.02475111186504364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,1024,0.012469333079126147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,768,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,768,0.023782221807373896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,512,0.00360622215602133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,768,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,512,0.011805333197116852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,256,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,512,0.024004444479942322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,256,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,128,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,1536,0.026134222745895386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,256,0.023003555006451074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,64,0.0032711111836963226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,128,0.011179555621412067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,2048,128,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,2048,32,0.003034666594531801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,65536,0.04862311151292589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,64,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,2048,32,0.010843555960390302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,51200,0.03962488969167074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,65536,0.14171377817789713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,51200,0.11075822512308757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,65536,0.2602710988786486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,16384,0.018636445204416912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,51200,0.20445777310265434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,12288,0.016552888684802584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,16384,0.07980177799860637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,16384,0.03984355595376756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,10240,0.015184000134468079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,12288,0.032263110081354775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,12288,0.06548000044292875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,8192,0.013751111096805997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,10240,0.05826666620042589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,10240,0.02829866607983907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,8192,0.024854222933451336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,7168,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,8192,0.050119111935297646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,7168,0.022787556052207947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,6144,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,7168,0.04801688922776116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,6144,0.020273778173658583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,5120,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,6144,0.04269422094027201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,5120,0.017948443690935772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,4096,0.008016000191370646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,5120,0.03772177630000644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,4096,0.016223111086421542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,3584,0.007032000356250339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,4096,0.0340577761332194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,3584,0.015832000308566622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,3072,0.007344889144102733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,3584,0.032481776343451604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,3072,0.014818666709793938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,2560,0.006643555644485686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,3072,0.031071109904183283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,2560,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,2048,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,2560,0.029170665476057265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,2048,0.013174222575293647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,1536,0.0055440002017551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,2048,0.027818666564093694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,1536,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,1536,0.025839999318122864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,1024,0.012506666282812754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,768,0.004231111043029361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,1024,0.025514667232831318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,768,0.0244168887535731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,768,0.012173333101802401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,512,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,512,0.01184177812602785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,1024,0.004547555413511065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,512,0.024097777075237695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,256,0.003506666670242945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,128,0.0029564443975687027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,256,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,256,0.02271288964483473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,128,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,64,0.0029360000044107437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1536,32,0.003089777711364958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,64,0.011518222590287527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,65536,0.039701332648595176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1536,32,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,65536,0.13979910479651556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,51200,0.03267911076545715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,65536,0.2603057755364312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,51200,0.10998666948742336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,16384,0.014639111028777229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,51200,0.20504355430603027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1536,128,0.018966222802797954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,12288,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,16384,0.03840800126393636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,16384,0.07957866456773546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,10240,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,12288,0.0652551121181912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,12288,0.032002667586008705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,10240,0.027848889430363972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,10240,0.05821955866283841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,8192,0.023461333579487268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,8192,0.04831822382079231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,7168,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,7168,0.0214684440029992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,7168,0.04695999953481886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,6144,0.010689777632554373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,6144,0.041504889726638794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,6144,0.020484444167878892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,5120,0.009370666411187913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,5120,0.0373564428753323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,8192,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,5120,0.017828444639841717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,4096,0.008007999923494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,3584,0.007296000089910295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,4096,0.03405244482888116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,3584,0.01551822159025404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,3072,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,3584,0.032305777072906494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,3072,0.014871999621391296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,2560,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,3072,0.030650665362675984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,2560,0.01388711151149538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,2048,0.005550222264395819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,2560,0.029214223225911457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,2048,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,1536,0.004924444274769889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,2048,0.027823110421498615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,1536,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,1536,0.025785777303907607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,1024,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,1024,0.02442311081621382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,1024,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,768,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,768,0.023353776997990076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,768,0.012142221960756512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,512,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,512,0.02269866731431749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,512,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,256,0.003273777870668305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,256,0.022333333889643352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,256,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,4096,0.01625955601533254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,128,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,1024,128,0.019912888606389362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,64,0.0031600000543726813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,64,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,1024,32,0.003048000029391713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,32,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,65536,0.02998577886157566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,1024,128,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,65536,0.13904533121320936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,51200,0.02741333345572154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,65536,0.2603057755364312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,51200,0.10905510849422878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,16384,0.013591110706329346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,51200,0.20478578408559164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,16384,0.03843999902407328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,12288,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,16384,0.07944711049397786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,12288,0.03192089001337687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,12288,0.06413955820931329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,10240,0.026690666874249775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,8192,0.009553777674833933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,8192,0.02308799988693661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,8192,0.048560000128216214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,7168,0.009045333498054082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,10240,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,7168,0.021685333715544805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,6144,0.010046222143703038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,7168,0.04493600130081177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,10240,0.05635199944178263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,5120,0.00906488878859414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,6144,0.041893333196640015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,6144,0.02014755540423923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,4096,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,5120,0.03739733497301737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,5120,0.017652443713612027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,4096,0.016326222154829238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,4096,0.03331999977429708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,3584,0.03159911102718777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,3584,0.015512888630231222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,3072,0.006351999938488007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,3072,0.030247999562157526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,3072,0.014815110299322339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,2560,0.005819555371999741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,2560,0.028492444091373022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,2560,0.013989332649442883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,2048,0.005160000175237656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,2048,0.02748888896571265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,2048,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,1536,0.004621333132187526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,3584,0.0069297779765393995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,1024,0.0042275554604000514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,1536,0.025753777888086107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,1536,0.012888000243239932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,768,0.0037991110649373797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,1024,0.024725332856178284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,1024,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,512,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,768,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,768,0.023787554767396715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,256,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,512,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,512,0.023708444502618577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,128,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,256,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,256,0.021964444054497614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,768,128,0.01891288823551602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,128,0.011494222614500256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,768,32,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,65536,0.024313777685165405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,32,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,768,64,0.011154666543006897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,51200,0.01943822205066681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,65536,0.13829688231150308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,65536,0.26014312108357746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,51200,0.2044844494925605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,16384,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,51200,0.10806400246090359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,12288,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,16384,0.03749777873357137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,16384,0.0795688894059923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,10240,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,12288,0.06465599934260051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,12288,0.03054666519165039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,8192,0.010206222534179688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,10240,0.057592001226213246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,10240,0.026416889495319788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,7168,0.00976533359951443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,8192,0.023090667194790308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,8192,0.04760444495413038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,6144,0.008661333057615492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,7168,0.02156177825397915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,6144,0.019912888606389362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,5120,0.008295999632941352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,6144,0.041535112592909075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,5120,0.017525333497259352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,4096,0.007627555893527136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,5120,0.03765777746836344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,4096,0.016207110550668504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,4096,0.033301333586374916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,3584,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,3584,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,7168,0.04535733328925239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,3072,0.0063777777055899305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,3584,0.03192089001337687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,3072,0.030233777231640283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,2560,0.005596444424655702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,3072,0.014859555496109856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,2560,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,2048,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,2560,0.02824977702564663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,2048,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,1536,0.004879111217127906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,2048,0.027496889233589172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,1536,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,1536,0.02548355526394314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,1024,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,1024,0.024393777052561443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,768,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,768,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,512,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,768,0.024137778414620295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,1024,0.0041946665280395085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,512,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,256,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,512,0.023754666248957317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,256,0.011374221907721626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,128,0.018922666708628338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,128,0.011381333072980246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,64,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,512,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,64,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,65536,0.01555999947918786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,512,32,0.010709332923094431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,65536,0.13698577880859375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,51200,0.0177502218219969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,65536,0.2605022324456109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,512,256,0.021670222282409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,16384,0.0118568887313207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,51200,0.1064515577422248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,51200,0.20497777726915148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,16384,0.03715466790729099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,12288,0.00998133338159985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,16384,0.07858844598134358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,10240,0.009009777671760982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,12288,0.029859556092156306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,12288,0.06263111035029094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,8192,0.008676444490750631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,10240,0.026152888933817547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,10240,0.055424888928731285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,7168,0.00798044436507755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,8192,0.022898667388492163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,8192,0.047320889102088086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,6144,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,7168,0.021310221817758348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,7168,0.044871111710866295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,5120,0.007709333466158972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,6144,0.04184622234768338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,5120,0.03704444567362467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,5120,0.017525333497259352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,4096,0.007287110719415877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,4096,0.03328266739845276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,4096,0.016232000456915963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,3584,0.006630222416586346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,3584,0.03165510959095425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,3584,0.01538222200340695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,3072,0.0059057776298787855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,3072,0.030165334542592365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,2560,0.005660444498062134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,6144,0.01991555591424306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,2560,0.013777777552604675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,2048,0.0052266667286554975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,2560,0.0284204449918535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,2048,0.013068444199032254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,2048,0.027429333991474573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,1536,0.01276000009642707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,1536,0.026068444053332012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,1024,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,3072,0.014464888307783337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,1024,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,768,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,1024,0.024456888437271118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,768,0.01200711147652732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,768,0.02307111190425025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,512,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,1536,0.004530666602982415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,256,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,512,0.022712000542216833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,512,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,128,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,256,0.022314666046036616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,256,128,0.019367999500698514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,64,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,128,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,256,32,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,64,0.011185777684052786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,65536,0.014841778410805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,32,0.010775999890433418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,65536,0.13466755549112955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,256,256,0.011374221907721626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,51200,0.013100444442696042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,65536,0.26011821958753795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,16384,0.009384888741705153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,51200,0.20474489529927573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,51200,0.10262488656573826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,12288,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,16384,0.07680533329645793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,16384,0.03638044330808852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,10240,0.008742222355471717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,12288,0.06235288911395603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,12288,0.03042844600147671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,8192,0.008070222205585903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,10240,0.05547555618815952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,10240,0.02605155607064565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,7168,0.007699555820888943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,8192,0.022895110978020564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,8192,0.04692444536421034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,6144,0.007296889192528195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,7168,0.020991999242040846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,7168,0.04490310947100321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,6144,0.01960533360640208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,5120,0.008007999923494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,6144,0.04160088962978787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,5120,0.017400888933075797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,4096,0.007358222372002072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,5120,0.03765689002143012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,3584,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,4096,0.015848888291252982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,4096,0.03336800138155619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,3072,0.006764444626039929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,3584,0.03166666626930237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,3584,0.015504888362354703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,2560,0.0063528890411059065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,3072,0.030240890052583482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,3072,0.01422044469250573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,2048,0.005971555494599872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,2560,0.013835555149449242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,2560,0.02851111027929518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,1536,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,2048,0.012892444100644855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,2048,0.027801778581407335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,1536,0.012464000119103326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,1024,0.003927999900446998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,1536,0.02587022218439314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,1024,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,1024,0.025363556212849084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,768,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,512,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,768,0.011534222298198275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,768,0.024393777052561443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,256,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,512,0.023031110564867657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,512,0.011485333244005838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,128,0.002993777808215883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,256,0.021664000219768945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,256,0.011558222273985544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,64,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,224,128,128,0.018590221802393597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,128,32,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,128,0.010791999598344168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,65536,0.01369066701995002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,64,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,128,32,0.010410666465759277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,51200,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,16384,0.009106666677527958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,51200,0.10495911041895549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,12288,0.008017777568764156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,16384,0.036045332749684654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,10240,0.007684444387753804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,12288,0.029517332712809246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,8192,0.007032000356250339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,10240,0.026114667455355327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,7168,0.006919999917348226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,8192,0.02269688910908169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,6144,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,7168,0.02139288849300808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,6144,0.0196142229768965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,5120,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,4096,0.006467555546098285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,5120,0.017349332571029663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,65536,0.13548889425065783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,3584,0.00630666646692488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,3072,0.006082666830884085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,4096,0.016210666961140104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,2560,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,3072,0.014531556102964612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,3584,0.015182221929232279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,2048,0.006272000157170826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,1536,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,2560,0.013815110756291283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,1024,0.003855111284388436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,2048,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,1536,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,768,0.0032995556377702286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,512,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,1024,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,256,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,768,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,512,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,256,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,64,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,128,0.01074399964676963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,64,32,0.002521777732504739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,64,0.010804444551467896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,64,32,0.01037599974208408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,65536,0.013360000318951078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,51200,0.01181244436237547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,65536,0.13443111048804388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,16384,0.008775110873911116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,51200,0.10417244169447158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,16384,0.03567022085189819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,12288,0.007716444631417592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,10240,0.007320889168315464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,12288,0.029522667328516643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,8192,0.007010666860474481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,10240,0.026120889517996047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,8192,0.022673777408070032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,7168,0.00665244460105896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,6144,0.006670222100284364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,7168,0.021031111478805542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,5120,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,6144,0.019613333874278598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,4096,0.005939555664857228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,5120,0.017250667015711468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,3584,0.006015111174848344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,4096,0.015973332855436537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,3584,0.01517599986659156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,3072,0.006007110906971826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,2560,0.006243555496136348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,3072,0.014466666513019137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,2048,0.005864000154866113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,2560,0.013602666556835175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,2048,0.012917333179050021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,1536,0.00444444434510337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,1024,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,1536,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,768,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,1024,0.011830222275522021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,768,0.01148888882663515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,512,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,256,0.002975111206372579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,512,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,128,0.002591111179855135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,256,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,128,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,64,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,224,32,32,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,64,0.010696888797812991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,224,32,32,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,16384,0.38741599188910586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,51200,1.1659919950697157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,16384,0.3352479934692383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,51200,1.0533520380655925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,12288,0.29208532969156903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,10240,0.24635643429226348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,12288,0.25286756621466744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,8192,0.20407466093699136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,10240,0.21237333615620932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,8192,0.17214400238460967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,7168,0.17335289054446748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,6144,0.15273511409759521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,7168,0.15394311481051975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,5120,0.1274968915515476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,5120,0.11251555548773871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,4096,0.10260533624225193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,6144,0.1324497726228502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,3584,0.09055378039677937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,4096,0.0931546688079834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,3072,0.07877688937717013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,2560,0.06723288695017497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,3072,0.073125335905287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,3584,0.08363199896282619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,2048,0.062421335114373096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,1536,0.04706044329537285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,2560,0.06438044706980388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,2048,0.05520533190833198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,1024,0.032348444064458214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,1536,0.045328001181284584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,768,0.024907555845048692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,512,0.018941332896550495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,768,0.031488001346588135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,1024,0.03715199894375271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,256,0.013373333546850415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,512,0.03165955675972833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,128,0.010185778141021729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,64,0.007961778177155389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,128,0.023209777143266466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,256,0.024481778343518574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,64,0.022721777359644573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,65536,32,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,65536,32,0.023471110396915015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,65536,1.0427689022488065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,65536,1.242884423997667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,51200,0.9758986367119683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,16384,0.3280888928307427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,12288,0.24352622032165527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,51200,0.830952008565267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,16384,0.26931556065877277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,10240,0.18755200174119738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,12288,0.203111105495029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,10240,0.17232710785335967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,8192,0.15531110763549805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,7168,0.1318168905046251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,8192,0.1400337749057346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,7168,0.12481421894497341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,6144,0.11611378192901611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,5120,0.09755377637015449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,6144,0.10800088776482476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,4096,0.07928711175918579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,5120,0.0922408898671468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,4096,0.07634933127297296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,3584,0.07012799713346693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,3072,0.06083466609319051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,3584,0.06859555509355333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,2560,0.051870223548677236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,3072,0.06127822399139404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,2560,0.05373422304789225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,2048,0.04323822259902954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,1536,0.035820444424947105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,2048,0.046147555112838745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,1024,0.025386666258176167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,1536,0.038904001315434776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,1024,0.03153866529464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,768,0.020375111036830477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,512,0.014731556177139282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,768,0.027243556247817144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,256,0.010031999813185798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,512,0.02757599949836731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,128,0.008994667066468133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,256,0.02188088827662998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,64,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,64,0.020478222105238173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,128,0.020322667227851022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,51200,32,0.007640000018808577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,51200,32,0.020790222618314955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,65536,0.36755021413167316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,51200,0.28908801078796387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,65536,0.3996968799167209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,16384,0.10141866736941868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,51200,0.31697511672973633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,12288,0.07500444518195258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,16384,0.10787910885281032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,10240,0.06383377975887723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,12288,0.08329066965315077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,8192,0.054323554039001465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,10240,0.07076266739103529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,8192,0.06005599763658312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,7168,0.04570400052600437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,6144,0.040911999013688825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,7168,0.05418666534953647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,5120,0.03448266784350077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,5120,0.04020444552103678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,6144,0.04603999853134155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,4096,0.028480887413024902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,3584,0.025421332981851365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,3584,0.032223112053341336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,3072,0.022740445203251306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,3072,0.028152889675564233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,2560,0.01979022224744161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,2560,0.025676444172859192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,2048,0.016192888220151264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,2048,0.022971555590629578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,1536,0.013214222258991666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,4096,0.03488355543878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,1536,0.02071644365787506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,1024,0.010123555858929953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,768,0.00886933339966668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,1024,0.01808355583084954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,512,0.006812444577614467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,768,0.01701955497264862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,256,0.005290666802061929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,512,0.01689511040846507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,128,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,128,0.013976888524161445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,64,0.0041751112374994485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,16384,32,0.00424888895617591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,64,0.014150222142537435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,32,0.01514311134815216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,65536,0.2854168944888645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,51200,0.22433598836263022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,65536,0.34265955289204914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,16384,0.08061155345704821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,16384,256,0.014602666099866232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,51200,0.2706711027357313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,12288,0.05899022022883097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,16384,0.0943031112353007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,10240,0.05085866649945577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,8192,0.04295911060439216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,10240,0.06289244360393949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,7168,0.036700444089041814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,8192,0.053245332505967885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,7168,0.04863022102249993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,6144,0.03275466627544827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,5120,0.028110222684012517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,6144,0.041179554329978094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,4096,0.02349866595533159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,5120,0.036093334356943764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,12288,0.07311199771033393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,4096,0.030841777722040813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,3584,0.020741333564122517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,3072,0.01796711153454251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,3072,0.02567377852069007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,3584,0.02824444572130839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,2560,0.015623999966515435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,2048,0.013608888619475894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,1536,0.011223110887739392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,2048,0.020911999874644812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,1024,0.009043555292818282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,2560,0.023578666978412207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,768,0.008364444805516137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,1536,0.01905866629547543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,512,0.005930666708283954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,1024,0.016560888952679105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,256,0.004902222090297275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,768,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,512,0.015558222929636637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,256,0.013815999858909182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,128,0.004268444246715969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,64,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,12288,32,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,128,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,64,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,12288,32,0.014217777384652032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,65536,0.23364443249172637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,51200,0.1838364468680488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,16384,0.06636622216966417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,51200,0.23929866154988608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,65536,0.3022186756134033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,16384,0.08398222261004978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,10240,0.04270844327078926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,12288,0.06575555271572538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,10240,0.05594222413169014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,8192,0.037793778710895114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,7168,0.03244977858331468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,8192,0.048327998982535474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,6144,0.0277137772904502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,7168,0.04380444354481167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,6144,0.03699377841419644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,5120,0.025285333395004272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,12288,0.05411110983954536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,4096,0.02212000058756934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,5120,0.03220444586541917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,4096,0.027775999572541978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,3584,0.02574133376280467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,3072,0.016695110334290397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,2560,0.01554666625128852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,3072,0.023751111494170293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,2048,0.012246222131782107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,2560,0.02201955517133077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,1536,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,2048,0.01997155613369412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,1024,0.00849422232972251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,1536,0.018171555466122098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,768,0.007689778175618913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,1024,0.015743111570676167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,512,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,768,0.014816888504558139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,256,0.004628444297446145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,3584,0.01991555591424306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,128,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,512,0.01481244464715322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,256,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,64,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,10240,32,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,128,0.012799999780125089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,64,0.012466666599114736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,10240,32,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,65536,0.21278756194644502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,51200,0.16956533326043022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,65536,0.258246209886339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,16384,0.07354844278759427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,12288,0.04426133301522997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,51200,0.2039804458618164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,10240,0.03853955533769395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,12288,0.05760800176196628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,8192,0.03161688976817661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,10240,0.050471110476387866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,8192,0.042066666815016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,7168,0.027777777777777776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,6144,0.024854222933451336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,7168,0.0377111103799608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,5120,0.0211173329088423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,6144,0.03233511249224345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,5120,0.028773334291246202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,4096,0.017342221405771043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,3584,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,4096,0.025032000409232244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,3072,0.013739555246300168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,3584,0.023726221587922838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,3072,0.021887110339270696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,2560,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,2048,0.010692444112565784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,2560,0.02029333346419864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,1536,0.008861333131790161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,2048,0.01829777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,1024,0.007221333682537079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,16384,0.05975288814968533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,1536,0.01684266659948561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,768,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,1024,0.01460088955031501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,512,0.004531555705600315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,512,0.014096889230940076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,256,0.003855111284388436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,256,0.012734221915404001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,128,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,128,0.012552888856993781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,64,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,64,0.01275644451379776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,8192,32,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,768,0.014115555418862237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,8192,32,0.013432000246312885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,65536,0.2018773290846083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,51200,0.15839821762508816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,65536,0.2228764428032769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,16384,0.056201775868733726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,51200,0.17757511138916016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,16384,0.06433333290947808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,12288,0.04222311245070564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,10240,0.03636711173587375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,12288,0.05053155620892843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,10240,0.04418933391571045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,8192,0.030412445465723675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,7168,0.026535110341178045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,8192,0.03668177790111966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,6144,0.02360533343421088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,7168,0.03307644526163737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,5120,0.019655999210145738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,6144,0.028423110644022625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,4096,0.016153777639071148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,5120,0.025285333395004272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,3584,0.014859555496109856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,4096,0.02268444498380025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,3072,0.01310488912794325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,3584,0.021410667233996924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,2560,0.011715555356608497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,3072,0.0199644449684355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,2048,0.01034400032626258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,2560,0.01889866590499878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,1536,0.010100444157918295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,2048,0.017162667380438913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,1024,0.007592889169851939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,1536,0.015456888410780164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,768,0.006252444452709622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,1024,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,512,0.005274666680230035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,768,0.013817778064144982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,256,0.004805333498451444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,256,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,128,0.004119111018048393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,64,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,128,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,7168,32,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,32,0.012875555290116204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,64,0.012710221939616732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,65536,0.17377066612243652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,51200,0.13708621925777859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,7168,512,0.013764444324705334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,16384,0.04805422160360548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,51200,0.17266044351789686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,12288,0.036710222562154136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,65536,0.21907199753655326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,16384,0.06306933032141791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,10240,0.03185866607560052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,8192,0.02662311163213518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,12288,0.049264887968699135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,10240,0.04253066579500834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,7168,0.022971555590629578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,6144,0.020427554845809937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,7168,0.03233688738611009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,5120,0.017508443858888414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,6144,0.027781334188249376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,4096,0.014136888914638095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,8192,0.03583911061286926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,3584,0.012950222525331708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,5120,0.025045333637131587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,3072,0.0118568887313207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,4096,0.02254933284388648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,2560,0.010086221827401055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,3072,0.019723556107944913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,2048,0.00904444439543618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,2560,0.018552000323931377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,1536,0.007686222592989604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,3584,0.021870222356584337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,2048,0.016866667403115165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,1024,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,1536,0.015117333994971381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,1024,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,512,0.004548444516128964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,768,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,256,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,512,0.01348622226052814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,128,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,256,0.01239111108912362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,64,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,128,0.012039999994966718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,32,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,64,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,6144,32,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,6144,768,0.005109333329730564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,65536,0.16364622116088867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,51200,0.12915289402008057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,65536,0.21491200394100615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,51200,0.1704782247543335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,16384,0.040130668216281466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,12288,0.03157155712445577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,16384,0.06202666627036201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,10240,0.02494577732351091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,10240,0.04079466727044847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,12288,0.048192888498306274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,8192,0.03479199939303928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,7168,0.020021332634819877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,7168,0.03184266553984748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,6144,0.017504000001483493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,6144,0.02790133489502801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,8192,0.02071644365787506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,5120,0.01603822244538201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,4096,0.012601777911186218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,5120,0.024808888634045918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,3584,0.011054221954610614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,4096,0.02234933276971181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,3072,0.010758221977286868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,3584,0.0210524449745814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,2560,0.009576000273227692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,3072,0.019535111056433786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,2048,0.008364444805516137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,2560,0.01759466694460975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,1536,0.006955555743641323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,2048,0.016185777054892648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,1024,0.005624889085690181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,1536,0.01516888870133294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,768,0.00488711107108328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,768,0.013423999978436364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,1024,0.01389244364367591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,512,0.004361777669853634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,256,0.0038933331767717996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,512,0.013424889081054263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,256,0.012435555458068848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,128,0.0036159998012913596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,64,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,128,0.012347555822796293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,5120,32,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,64,0.012395555774370829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,5120,32,0.012455999851226807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,65536,0.17783199416266549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,51200,0.09427555402119954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,51200,0.14130400286780462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,16384,0.038329776790406965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,65536,0.11926133102840847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,16384,0.05160799953672621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,12288,0.028597331709331934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,10240,0.02463200026088291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,12288,0.04006933172543844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,8192,0.01999200052685208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,10240,0.03411111235618591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,8192,0.029847999413808186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,7168,0.01793688866827223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,6144,0.015670221712854173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,7168,0.027466666367318895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,5120,0.01402933398882548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,6144,0.024525332782003615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,5120,0.021864889396561518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,4096,0.014187556174066333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,3584,0.012855110896958245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,4096,0.020030222005314298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,3072,0.012852444416946836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,3584,0.019133332702848647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,2560,0.010070222119490305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,3072,0.017272000511487324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,2048,0.009407111340098912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,2560,0.016144000821643405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,2048,0.014911999305089315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,1536,0.0069653333889113525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,1024,0.006333333336644703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,1536,0.014137778017255994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,768,0.0053297777970631914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,1024,0.013455111119482251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,512,0.004610666798220741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,768,0.01309511148267322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,256,0.003996444245179494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,512,0.012934221989578672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,256,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,128,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,64,0.0032586666444937387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,64,0.01202222208182017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,32,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,4096,32,0.012333333492279053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,65536,0.11209955480363633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,4096,128,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,51200,0.08881244394514297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,65536,0.1746008925967746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,16384,0.03625599874390496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,16384,0.050123555792702564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,12288,0.028035554620954726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,12288,0.039136889908048846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,10240,0.02609777781698439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,10240,0.03372888763745626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,8192,0.025495999389224585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,51200,0.13825066884358725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,8192,0.029312888781229656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,7168,0.0203448881705602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,6144,0.01805688937505086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,7168,0.0273973329199685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,5120,0.017616000440385606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,6144,0.024077332682079736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,5120,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,4096,0.013864888913101621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,3584,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,4096,0.019461333751678467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,3072,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,3584,0.018858666221300762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,3072,0.0163111107216941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,2560,0.015660444895426433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,2048,0.008640888664457533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,2048,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,1536,0.007639110916190677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,1536,0.013823111024167804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,1024,0.006317333214812809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,1024,0.012871999707486896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,2560,0.00942844483587477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,768,0.00462666650613149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,512,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,512,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,768,0.013148444394270578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,256,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,128,0.00327822214199437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,128,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,256,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,64,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3584,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,32,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3584,64,0.0120666664507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,65536,0.08624533149931167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,51200,0.06993689139684041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,65536,0.1711724466747708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,16384,0.029701332251230877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,51200,0.13629778226216635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,12288,0.02387822171052297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,16384,0.048932443062464394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,12288,0.037626667155159846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,10240,0.022797332869635686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,8192,0.017479111750920612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,10240,0.03314400050375197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,7168,0.01664444473054674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,8192,0.02922399838765462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,6144,0.012905778156386482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,7168,0.02682755556371477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,5120,0.011407111254003314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,6144,0.023826667004161414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,5120,0.021531555387708876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,4096,0.01127911110719045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,3584,0.010452444354693094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,4096,0.018385777870814007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,3072,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,3584,0.01831644442346361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,2560,0.008301333420806462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,3072,0.016162667009565566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,2048,0.006744888921578725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,2560,0.015923556354310777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,1536,0.005984000033802456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,2048,0.014452444182501899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,1024,0.0046951112647851305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,1536,0.014192889134089151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,768,0.004553777890072929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,1024,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,512,0.003916444463862313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,768,0.013065777719020844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,512,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,256,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,128,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,256,0.011782222323947482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,64,0.0031395554542541504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,128,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,3072,32,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,64,0.011062222222487131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,65536,0.07925244172414143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,3072,32,0.01237511138121287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,51200,0.06419466601477729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,65536,0.16972000069088408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,16384,0.027033777700530157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,51200,0.13524444897969565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,12288,0.021108445194032457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,16384,0.04801155461205376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,10240,0.018736888964970905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,8192,0.017638223038779367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,12288,0.0374017788304223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,7168,0.012835555606418185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,8192,0.028576887316173975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,6144,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,10240,0.032999111546410456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,7168,0.026836444934209187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,5120,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,6144,0.02329155637158288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,4096,0.010326222413116032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,3584,0.010079999764760336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,5120,0.020995555652512446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,4096,0.01789688898457421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,3072,0.009347555538018545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,3584,0.01716711123784383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,2560,0.008024888734022776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,3072,0.01645866698688931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,2048,0.006664888726340399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,2560,0.015213333898120456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,2048,0.01462222304609087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,1536,0.005694222119119432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,1024,0.004606222112973531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,1536,0.014088888963063559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,768,0.004476444588767158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,1024,0.012803555362754397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,512,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,768,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,256,0.0035582222044467926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,512,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,128,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,256,0.01203733351495531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,64,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,128,0.011036444041464063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2560,32,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,64,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,65536,0.06621688604354858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2560,32,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,51200,0.05189333359400431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,65536,0.13124710983700222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,16384,0.022272888157102797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,51200,0.10534755388895671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,12288,0.0192577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,16384,0.03839288817511664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,10240,0.01759733259677887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,12288,0.030039111773173015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,8192,0.01351199961370892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,10240,0.027162666122118633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,7168,0.012468444804350534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,8192,0.023984889189402264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,6144,0.011547555526097616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,5120,0.00979022185007731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,7168,0.022804444034894306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,4096,0.00868977771864997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,5120,0.017655111021465726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,3584,0.008198222352398766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,4096,0.016306666864289176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,3072,0.007996444072988298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,3584,0.015600000818570455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,2560,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,3072,0.014569777581426831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,2048,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,2560,0.014295111099878946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,1536,0.0052826665341854095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,2048,0.013532444834709167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,1024,0.004276444514592488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,1536,0.013177778157922955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,6144,0.018911111685964797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,768,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,512,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,1024,0.012475555141766867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,256,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,768,0.012418666647540199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,128,0.003157333367400699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,512,0.01216533366176817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,64,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,256,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,128,0.010777778095669217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,2048,32,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,64,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,65536,0.049941334459516734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,2048,32,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,65536,0.12875821855333117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,16384,0.017895110779338412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,51200,0.10312088992860581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,12288,0.015947555502255756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,16384,0.0370435549153222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,10240,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,12288,0.029765334394243028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,10240,0.02685955497953627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,8192,0.01311644415060679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,7168,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,8192,0.02350666622320811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,6144,0.010560000108347999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,51200,0.04086844457520379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,7168,0.021552888883484736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,5120,0.009655999640623728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,6144,0.018671999375025432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,5120,0.016964443855815463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,4096,0.015821332732836407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,3584,0.007496000164084964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,3072,0.006704000135262807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,3584,0.01519377695189582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,2560,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,3072,0.014282666974597506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,2048,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,2560,0.01402666668097178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,1536,0.004967111266321606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,1536,0.012868444124857584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,2048,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,1024,0.004886222382386525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,4096,0.00832622249921163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,1024,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,512,0.0039813332259655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,768,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,512,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,256,0.0035866668654812705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,256,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,128,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,768,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,64,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,128,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1536,32,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,64,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,65536,0.035951998498704694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1536,32,0.011644444531864591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,51200,0.030215998490651447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,65536,0.1259102159076267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,16384,0.014228444960382251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,51200,0.1001146634419759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,16384,0.03625511129697164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,12288,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,10240,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,12288,0.029994666576385498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,8192,0.011311111350854238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,10240,0.026110221942265827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,7168,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,8192,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,6144,0.010232888989978367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,7168,0.02069244450993008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,5120,0.010435555544164447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,6144,0.018604444132910836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,4096,0.007997333175606197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,5120,0.017151110702090792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,4096,0.01603200038274129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,3584,0.0073315559162033936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,3072,0.008031999899281396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,3584,0.015205333630243937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,2560,0.005976000179847081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,3072,0.014147554834683737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,2048,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,2560,0.013822221921549903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,1536,0.004635555462704765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,2048,0.013438222308953604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,1024,0.004257777912749184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,1536,0.013074666261672974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,768,0.004039999925427967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,512,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,1024,0.012319111161761813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,768,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,256,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,128,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,512,0.011888888974984487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,256,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,64,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,1024,32,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,128,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,64,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,65536,0.030127998855378892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,1024,32,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,51200,0.027406222290462915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,16384,0.013646221823162503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,65536,0.12484178278181289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,12288,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,51200,0.09874400165345933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,10240,0.010747555229398938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,12288,0.0294773346847958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,8192,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,8192,0.02200800014866723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,7168,0.00902133352226681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,7168,0.020598222812016804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,6144,0.010021333065297868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,6144,0.018548443913459778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,5120,0.008338666624493068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,16384,0.03602755400869582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,5120,0.017128000656763714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,4096,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,3584,0.007364444434642792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,10240,0.024736000431908503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,3072,0.00665866666369968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,4096,0.016207999653286405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,2560,0.006079999936951532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,3072,0.014135110709402295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,2048,0.005362666729423735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,3584,0.015303111738628812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,2560,0.013850666582584381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,1536,0.0047484445903036325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,2048,0.013533333109484779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,1536,0.013014222184816996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,1024,0.004265777766704559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,768,0.004120888809363048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,512,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,1024,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,256,0.0034204444123639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,512,0.01221777747074763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,768,0.012542222109105853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,256,0.011680888632933298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,64,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,768,32,0.002961777771512667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,64,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,65536,0.02358666724628872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,32,0.01145866678820716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,51200,0.022355554832352534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,65536,0.12424266338348389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,51200,0.09799377785788642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,16384,0.01111644423670239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,768,128,0.011536888778209686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,16384,0.03561955690383911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,12288,0.012509333590666452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,10240,0.01275288893116845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,12288,0.027514666318893433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,8192,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,10240,0.02446844511561924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,7168,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,8192,0.02186222208870782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,6144,0.009690666364298927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,7168,0.020609777834680345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,5120,0.008654221892356873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,6144,0.018473777506086562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,5120,0.01692800058258904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,4096,0.015850666496488783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,3584,0.006963555183675554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,3584,0.01519377695189582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,3072,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,3072,0.01421688828203413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,2560,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,4096,0.007351111206743453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,2560,0.013806222213639153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,2048,0.013090666797426013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,1536,0.004552000098758274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,1536,0.01274222218328052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,1024,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,1024,0.012110222544935016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,768,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,2048,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,512,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,768,0.012088889049159156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,256,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,512,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,256,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,128,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,64,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,128,0.01111644423670239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,512,32,0.0029315555261241067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,64,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,65536,0.016176000237464905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,512,32,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,51200,0.015327110886573792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,65536,0.1230613390604655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,16384,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,51200,0.09690844350390965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,12288,0.009381333159075843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,16384,0.03367822368939718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,10240,0.009054222040706212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,12288,0.02774222195148468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,8192,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,10240,0.024572443630960252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,7168,0.009006222089131674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,8192,0.02163999941613939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,6144,0.009550222092204625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,7168,0.020391111572583515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,5120,0.008293333152929941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,6144,0.0184551113181644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,4096,0.007295110987292395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,5120,0.016885333591037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,3584,0.006603555546866522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,4096,0.015628443823920358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,3072,0.005946666830115848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,3584,0.014860444598727755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,2560,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,3072,0.014109333356221518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,2048,0.00490399988161193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,2560,0.01351111133893331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,1536,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,2048,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,1024,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,1536,0.012680889003806643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,768,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,1024,0.012110222544935016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,512,0.003265777809752358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,768,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,256,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,512,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,128,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,256,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,64,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,128,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,256,32,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,64,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,256,32,0.011148444480366178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,51200,0.012863111164834766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,65536,0.12097244792514378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,16384,0.009389333426952362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,51200,0.09425955348544651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,12288,0.009357333183288574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,16384,0.03348888953526815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,10240,0.008779555559158325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,12288,0.027456889549891155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,8192,0.007981333467695449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,10240,0.024364444944593642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,7168,0.007603555917739868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,8192,0.021730666359265644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,6144,0.008015111088752747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,7168,0.020281778441535104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,65536,0.015192889504962497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,5120,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,6144,0.018248000078731112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,4096,0.00701422244310379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,4096,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,5120,0.01663644446267022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,3584,0.006597333484225803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,3072,0.005959110955397288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,3072,0.013845333622561561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,3584,0.014522666732470194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,2560,0.00554933316177792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,2048,0.005218666460778978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,2048,0.012851555314328937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,2560,0.013185777597957186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,1536,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,1024,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,1024,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,1536,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,768,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,512,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,512,0.011506666739781698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,768,0.012095111111799875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,256,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,128,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,256,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,128,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,64,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,128,32,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,64,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,65536,0.01348888874053955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,128,32,0.010436444646782346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,51200,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,65536,0.1209004455142551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,16384,0.0090506664580769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,51200,0.09513511260350545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,16384,0.03344622254371643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,12288,0.00942666663063897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,10240,0.008496888809733922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,12288,0.0273991111252043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,8192,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,7168,0.008664000365469191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,10240,0.024306666519906785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,6144,0.008025777836640676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,7168,0.02029244436158074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,5120,0.007290666302045186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,8192,0.02181333303451538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,6144,0.01797422269980113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,4096,0.007043555378913879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,3584,0.00664444433318244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,4096,0.015172445111804538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,3072,0.006051555689838197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,3584,0.015191111299726697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,2560,0.005744888964626525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,3072,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,2048,0.004929777648713854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,2560,0.013748444616794586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,1536,0.004586666822433472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,2048,0.012727999852763282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,1024,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,1536,0.01275644451379776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,5120,0.016639999217457242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,768,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,512,0.003293333368168937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,1024,0.011854222251309289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,256,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,512,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,128,0.002946666752298673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,256,0.011217777927716574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,64,0.002586666701568498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,128,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,64,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,64,0.010423999693658618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,32,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,65536,0.022696000006463792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,51200,0.0148044443792767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,65536,0.12033955256144206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,64,768,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,16384,0.009975999593734741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,51200,0.0945697757932875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,12288,0.009371555513805812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,10240,0.008737777670224508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,16384,0.03351199958059523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,8192,0.007672888537247975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,12288,0.02711822258101569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,7168,0.00871377769443724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,10240,0.024676443801985845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,6144,0.00833511104186376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,8192,0.02162488963868883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,5120,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,7168,0.019899555378490023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,6144,0.01846933364868164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,5120,0.01704088846842448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,4096,0.007383999725182851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,3584,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,3072,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,4096,0.015849777393870883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,2560,0.005606222069925732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,3072,0.013918222652541267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,3584,0.015072888798183866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,2048,0.00489155575633049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,2560,0.013813333378897773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,1536,0.004553777890072929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,2048,0.012827555338541666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,1024,0.003607999947335985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,1024,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,1536,0.01276622215906779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,768,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,512,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,768,0.012383999923865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,512,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,256,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,128,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,256,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,64,0.0028204443968004654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,128,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,193,32,32,0.0025822222232818604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,64,0.010778666370444827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,193,32,32,0.011367999845080905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,51200,1.164878209431966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,51200,0.8635546896192762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,16384,0.39321067598130965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,16384,0.2795333332485623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,12288,0.2905840078989665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,16384,0.28399822447035045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,12288,0.2116079992718167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,10240,0.2233955595228407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,10240,0.1785351170433892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,12288,0.2192080020904541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,8192,0.1839617755677965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,10240,0.189665781127082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,8192,0.144305772251553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,7168,0.15763821866777208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,51200,0.8351502418518066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,7168,0.128565337922838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,8192,0.15531199508243138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,6144,0.13949867089589438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,6144,0.11078488826751709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,7168,0.1405120028389825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,5120,0.11597155200110541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,5120,0.09424355294969346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,4096,0.09542933437559341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,6144,0.12277066707611084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,5120,0.10728266504075791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,4096,0.07814577553007337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,3584,0.08370400137371486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,4096,0.09196266863081191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,3072,0.07257599963082208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,3584,0.08359289169311523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,3072,0.06301422251595391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,2560,0.06233421961466471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,3072,0.07618311378690931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,2560,0.05530044436454773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,2048,0.05160089002715217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,2560,0.06900355550977919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,2048,0.047616001632478505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,1536,0.03959466682540046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,2048,0.061312887403700084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,1536,0.039832890033721924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,1024,0.028148445818159316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,1536,0.05439822210205925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,1024,0.03270666797955831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,1024,0.0461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,768,0.023207111491097346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,768,0.027686221732033625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,512,0.01646488904953003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,768,0.04271022147602505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,256,0.013615111509958902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,512,0.03885422150293986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,256,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,512,0.023035556077957153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,128,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,256,0.0343431101904975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,64,0.009635555247465769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,65536,128,0.029695110188590154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,128,0.019714666737450492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,65536,32,0.011870221959220039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,64,0.020277332928445604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,3584,0.07014933559629652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,65536,32,0.026750221848487854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,65536,0.8800480100843641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,65536,1.1206844117906358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,51200,0.8804648717244467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,16384,0.3020728958977593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,51200,0.696291552649604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,16384,0.22594756550259062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,12288,0.2208933300442166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,12288,0.16989244355095756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,16384,0.24886043866475424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,10240,0.18434666262732613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,10240,0.14395822419060603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,12288,0.19213422139485678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,8192,0.1529359949959649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,8192,0.1170097721947564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,10240,0.1647146675321791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,7168,0.12983732753329807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,7168,0.10448622041278416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,8192,0.13744266827901205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,6144,0.11442932817671035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,51200,0.7292248937818738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,65536,0.9228870603773328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,7168,0.12349511517418756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,6144,0.08977066808276707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,5120,0.09522844685448541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,5120,0.07668888568878174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,6144,0.10912177960077922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,4096,0.0776711106300354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,5120,0.09605599774254693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,3584,0.06855199734369914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,4096,0.06466577450434367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,3584,0.05765066544214884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,4096,0.08180444770389132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,3584,0.07548977931340535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,3072,0.05963289075427585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,2560,0.05115288827154371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,3072,0.05195200112130907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,2560,0.045929776297675244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,3072,0.06886666350894503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,2048,0.04180711176660326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,2048,0.040170666244294904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,2048,0.05632444222768148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,1536,0.032497776879204646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,2560,0.06263466676076253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,1536,0.03366488880581326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,1536,0.048784001006020435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,1024,0.027449778384632532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,768,0.020071110791630216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,1024,0.043325334787368774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,768,0.02316444449954563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,512,0.014836443795098199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,768,0.03913333349757724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,512,0.020447111792034574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,512,0.0350524451997545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,256,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,256,0.03232977787653605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,1024,0.02555111050605774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,256,0.01926933394538032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,128,0.00850133349498113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,64,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,51200,128,0.02571288910177019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,51200,32,0.006481777876615524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,128,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,32,0.01862044466866387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,65536,0.37100709809197324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,65536,0.363554663128323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,51200,64,0.018248889181349013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,51200,0.2881351047092014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,51200,0.2916533417171902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,16384,0.10203733046849568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,16384,0.09866222408082749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,51200,0.30945777893066406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,12288,0.07492888636059232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,16384,0.11537778377532959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,65536,0.3920168876647949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,10240,0.06393600172466703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,12288,0.09242133299509685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,10240,0.06526044342252943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,8192,0.05336088935534159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,10240,0.08124533626768324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,8192,0.05513155460357666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,7168,0.04617777797910902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,8192,0.0702622201707628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,12288,0.07659910784827338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,6144,0.04044355617629157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,7168,0.05036711030536228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,7168,0.06509866979387072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,6144,0.04226311047871908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,5120,0.03462933169470893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,6144,0.05846755372153389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,5120,0.03764444589614868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,4096,0.02884888980123732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,5120,0.053117334842681885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,3584,0.025518221987618342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,4096,0.03229688935809665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,4096,0.047269334395726524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,3072,0.022966222630606756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,3584,0.029662221670150757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,3584,0.04471733172734579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,2560,0.0196079992585712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,3072,0.026382222771644592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,3072,0.04170755545298258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,2048,0.016339555382728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,2560,0.023927999867333308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,2560,0.03916800022125244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,1536,0.013262222210566202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,2048,0.021331555313534204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,2048,0.036481777826944985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,1024,0.010440888504187265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,1536,0.03295822276009454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,1536,0.01923644377125634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,768,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,1024,0.01722666621208191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,1024,0.031222220924165513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,512,0.0076142218377855085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,768,0.015127999915017022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,512,0.014469333820872836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,512,0.02575911084810893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,256,0.005220444252093633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,256,0.024001777172088623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,256,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,128,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,128,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,128,0.02094399929046631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,64,0.004560889055331548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,16384,768,0.028320001231299505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,16384,32,0.004535111288229625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,64,0.013088000317414602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,16384,32,0.014101333088344999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,65536,0.273272885216607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,65536,0.28073867162068683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,51200,0.21336000495486787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,51200,0.22333688206142852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,65536,0.38757599724663633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,16384,0.07842399676640828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,16384,0.07875111367967394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,12288,0.05638400051328871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,16384,0.11547288629743789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,51200,0.30747466617160374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,12288,0.06173955731921726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,10240,0.04775466521581014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,10240,0.05270311236381531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,8192,0.04102399945259094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,12288,0.09217066897286309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,10240,0.08075822061962552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,7168,0.034671111239327326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,8192,0.045475555790795215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,7168,0.04175644450717502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,8192,0.0706604454252455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,6144,0.031018667750888403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,7168,0.06495466497209337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,5120,0.026497777965333726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,6144,0.05784710910585192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,6144,0.035092443227767944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,5120,0.030571556753582422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,4096,0.022081777453422546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,5120,0.05293155709902445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,3584,0.01955200069480472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,4096,0.02604533400800493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,4096,0.04756089051564535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,3072,0.01718133356836107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,3584,0.024298666252030268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,3584,0.044080889887279935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,2560,0.01478488908873664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,3072,0.02269777821169959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,3072,0.04117777612474229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,2048,0.012746666868527731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,2560,0.02075111038155026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,2048,0.03575911124547323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,2048,0.018971555762820773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,1536,0.010417777631017896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,1536,0.01756177842617035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,1536,0.032636443773905434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,1024,0.00871733327706655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,1024,0.015625778171751235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,1024,0.02970222300953335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,768,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,768,0.014488889111412896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,512,0.006008000009589725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,512,0.01349511080318027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,2560,0.03965955641534593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,256,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,512,0.026429333620601233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,256,0.024134222004148696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,256,0.013186666700575085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,128,0.004344888859324985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,128,0.021324444148275588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,128,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,64,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,12288,32,0.003620444486538569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,12288,768,0.027121777335802715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,64,0.013074666261672974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,12288,32,0.014456000592973499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,65536,0.28170400195651585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,51200,0.2318293253580729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,65536,0.26947376463148326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,51200,0.2130746708975898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,16384,0.07641688982645671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,16384,0.07636000050438775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,65536,0.3879866600036621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,12288,0.05372533202171326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,51200,0.3062640031178792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,16384,0.11450666851467556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,10240,0.049024889866511025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,12288,0.05974666939841377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,12288,0.09178666936026679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,10240,0.051318221622043185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,8192,0.03807466559939914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,10240,0.08044089211357965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,7168,0.03303466571701898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,8192,0.044157332844204374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,8192,0.0701120032204522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,7168,0.039882666534847684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,6144,0.029440889755884807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,7168,0.06438310941060384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,5120,0.02732177740997738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,6144,0.03366044494840834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,6144,0.057908442285325795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,5120,0.029159112109078303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,4096,0.020920000142521326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,5120,0.052637332015567355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,4096,0.025402666793929204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,3584,0.020290666156344943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,4096,0.046712001164754234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,3584,0.023759111762046814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,3584,0.04400266541375054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,3072,0.022348443667093914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,3072,0.04142133394877116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,2560,0.014893333117167154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,2560,0.0207324441936281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,3072,0.01827111177974277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,2560,0.03867377837498983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,2048,0.01367022262679206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,2048,0.035255110926098294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,2048,0.018910222583346896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,1536,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,1536,0.032038221756617226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,1536,0.017303110824690927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,1024,0.008033778104517195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,1024,0.02825511164135403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,1024,0.015278221832381355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,768,0.0069919998447100324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,768,0.026506667335828144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,768,0.014205333259370593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,512,0.005599110904667113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,512,0.025241777300834656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,512,0.01347555551264021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,256,0.004574222283230888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,256,0.023642667465739783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,256,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,128,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,10240,128,0.02068622244728936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,64,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,128,0.012737777498033313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,10240,32,0.0038613333470291565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,64,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,10240,32,0.013077333569526672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,65536,0.18973867098490396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,65536,0.22869955168830025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,51200,0.14944177203708225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,51200,0.18073422378963896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,16384,0.054647111230426364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,65536,0.260879119237264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,16384,0.06597155332565308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,51200,0.2088008854124281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,12288,0.03995733459790548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,16384,0.08504622512393528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,10240,0.03468888998031616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,12288,0.05161777800983853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,12288,0.06892800331115723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,10240,0.04526488979657491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,8192,0.029519110918045044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,10240,0.0613386697239346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,7168,0.025233777032958135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,8192,0.05373333228958977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,7168,0.033596442805396184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,6144,0.02271822260485755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,7168,0.05036711030536228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,6144,0.028912001185946997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,5120,0.019103111492262948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,6144,0.04491200049718221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,5120,0.025785777303907607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,5120,0.04186311033036974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,4096,0.016040000650617812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,8192,0.03779733180999756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,3584,0.01442577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,4096,0.02310400042268965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,4096,0.037215110328462385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,3072,0.012869333227475485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,3584,0.021418665846188862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,3072,0.02010222276051839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,3584,0.036462220880720354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,3072,0.0343413319852617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,2560,0.03256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,2048,0.009743111001120673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,2560,0.019117333822780185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,2048,0.017254221770498488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,1536,0.00831822223133511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,2048,0.031333333916134305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,1536,0.01556711064444648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,1024,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,1536,0.02899466620551215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,1024,0.01387644476360745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,2560,0.011318222516112857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,1024,0.02619022296534644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,768,0.013387555877367655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,512,0.00452533322903845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,512,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,768,0.025066667132907446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,256,0.003936888857020272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,512,0.023717333873112995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,256,0.022073777185546026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,256,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,128,0.004217777815130022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,768,0.005335999859703912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,64,0.003951999876234266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,8192,128,0.01923822197649214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,8192,32,0.0038737778862317405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,128,0.012322666744391123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,65536,0.17658400535583496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,32,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,65536,0.2230471240149604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,8192,64,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,51200,0.13931555218166775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,65536,0.257113774617513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,16384,0.05087466703520881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,51200,0.20551644431220162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,16384,0.06490133206049602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,12288,0.03740444448259141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,51200,0.1777200036578708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,12288,0.050959110260009766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,16384,0.08439911074108547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,10240,0.03242311212751601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,12288,0.06756622261471219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,10240,0.04492888847986857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,8192,0.027260444230503503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,10240,0.06001066499286228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,7168,0.02351288828584883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,8192,0.05244355400403341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,8192,0.03661244445376926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,7168,0.03332533439000448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,6144,0.021003555920388963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,7168,0.049161778555976025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,5120,0.017976888351970248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,6144,0.028129776318868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,6144,0.04439111219512092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,4096,0.014856000741322836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,5120,0.04086755712827047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,5120,0.025469332933425903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,3584,0.013638222383128272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,4096,0.03704444567362467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,4096,0.02291999922858344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,3072,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,3584,0.021342222889264423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,3584,0.03532622257868449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,2560,0.010619555910428366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,3072,0.020191111498408847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,3072,0.0336737765206231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,2560,0.01867733399073283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,2048,0.009396444592210982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,2560,0.03230133321550157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,1536,0.007958222594526079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,2048,0.016948444975747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,2048,0.030581331915325586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,1024,0.00665244460105896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,1536,0.027127110295825537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,1536,0.015599111715952555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,768,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,1024,0.02490666674243079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,1024,0.013888888888888888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,512,0.004977777600288391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,768,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,768,0.024162666665183172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,256,0.003951999876234266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,512,0.012818666795889536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,512,0.023727110690540735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,256,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,128,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,256,0.02236355510022905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,128,0.012126222252845764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,64,0.003313777761326896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,7168,128,0.02072355482313368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,7168,32,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,64,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,65536,0.17399822341071236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,65536,0.18692533175150552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,51200,0.13728977574242487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,51200,0.14855644438001844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,65536,0.24317333433363172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,16384,0.04807733496030172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,7168,32,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,16384,0.05531733234723409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,12288,0.03683733277850681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,51200,0.19473511642879912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,16384,0.07908711168501112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,10240,0.03170844581392076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,12288,0.06455910868114896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,12288,0.04368889000680712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,10240,0.03807199994723002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,8192,0.02644088864326477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,10240,0.05746489100986057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,7168,0.022924444741672937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,8192,0.03155288762516446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,8192,0.050854222642050854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,6144,0.020302222834693063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,7168,0.02867022156715393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,7168,0.04773155517048306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,5120,0.017285333739386666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,6144,0.02473066747188568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,6144,0.042698668109046094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,4096,0.014135999812020196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,5120,0.022590221630202398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,5120,0.039566222164365984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,4096,0.020454222957293194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,3584,0.01295555548535453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,4096,0.036791112687852644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,3584,0.019137778215938143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,3072,0.011596444580290051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,3584,0.03470044334729513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,2560,0.010138666464222802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,3072,0.018290667070282828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,3072,0.03375377919938829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,2560,0.03200177682770623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,2048,0.008791999684439765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,2048,0.015871110889646742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,2048,0.03001777662171258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,1536,0.007637333538797166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,1536,0.014363555444611443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,2560,0.017089777522616915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,1024,0.005647111270162795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,1536,0.02607022225856781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,1024,0.013167999684810638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,1024,0.025711999999152288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,768,0.004871110949251386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,768,0.02440888848569658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,768,0.013071999781661563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,512,0.004264889078007804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,512,0.023034666975339253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,512,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,256,0.012088889049159156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,256,0.02272266646226247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,128,0.003597333199448056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,128,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,256,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,64,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,6144,128,0.020302222834693063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,6144,32,0.003264888914095031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,64,0.011711999773979187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,6144,32,0.01219555570019616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,65536,0.14468888441721597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,65536,0.18194222450256348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,51200,0.11406755447387695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,65536,0.24108889367845324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,16384,0.04502400093608432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,51200,0.14633244938320583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,16384,0.0536480016178555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,12288,0.03454755412207709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,51200,0.1932213306427002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,12288,0.06383911106321546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,12288,0.0412124428484175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,10240,0.02996088729964362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,10240,0.03545244534810384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,10240,0.0569173362520006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,8192,0.02459022237194909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,16384,0.07812622520658705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,8192,0.05041955577002632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,7168,0.02165511084927453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,7168,0.02796088986926609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,6144,0.019102222389645047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,6144,0.024382222029897902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,6144,0.04232177800602383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,8192,0.030230220821168687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,5120,0.01632000009218852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,5120,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,5120,0.03973689013057285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,7168,0.04755555589993795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,4096,0.013467555244763693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,3584,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,4096,0.036002665758132935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,4096,0.01996799972322252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,3072,0.011519110865063138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,3584,0.018935999936527677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,3584,0.03460622164938185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,2560,0.01014311114947001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,3072,0.03255022234386868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,3072,0.01800622211562263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,2048,0.00884622252649731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,2560,0.016925333274735343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,2560,0.031239112218221027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,1536,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,2048,0.029707554313871596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,2048,0.014473777678277759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,1024,0.005394666559166378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,1536,0.013843555417325763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,1536,0.025721778472264607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,768,0.004679999831649992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,1024,0.02405333353413476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,1024,0.013146667016877068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,512,0.004974222017659081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,768,0.012770666844315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,768,0.023425777753194172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,256,0.003979555434650845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,512,0.012346666720178394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,512,0.023698666029506262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,128,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,256,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,256,0.022077333596017625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,64,0.003313777761326896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,5120,128,0.018608889646000333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,128,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,5120,32,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,65536,0.10683644480175442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,32,0.012141333685980903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,65536,0.1716302235921224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,51200,0.0838737752702501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,65536,0.24039199617173937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,51200,0.13813777764638266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,16384,0.03666844301753574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,5120,64,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,16384,0.0518142216735416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,12288,0.028505778974956934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,51200,0.19164978133307564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,16384,0.07772088713116117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,12288,0.039876444472206965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,10240,0.023806222611003455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,12288,0.06252088811662462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,10240,0.03414577907986111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,8192,0.019920888874265883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,10240,0.056367112530602344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,7168,0.018265777164035372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,8192,0.029493331909179688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,8192,0.05014933480156792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,6144,0.015541333291265698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,7168,0.02753066685464647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,7168,0.0469760000705719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,5120,0.013838222457302941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,6144,0.024248000648286607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,6144,0.04225422276390923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,4096,0.00978577799267239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,5120,0.03906933466593424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,5120,0.02203377750184801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,3584,0.009395555489593083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,4096,0.0358862214618259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,4096,0.01995644470055898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,3072,0.008362666600280339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,3584,0.01860000027550591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,3584,0.03401777810520596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,2560,0.007616889145639207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,3072,0.016910221841600206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,3072,0.0329039990901947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,2048,0.007338666253619724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,2560,0.016238222519556682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,2560,0.031088001198238794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,2048,0.02787644333309597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,2048,0.015071999695565967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,1536,0.005683555371231503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,1024,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,1536,0.013850666582584381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,1536,0.025819554924964905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,768,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,1024,0.012821333275900947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,1024,0.024672889047198828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,512,0.0036026665733920205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,768,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,512,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,256,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,512,0.023386667172114056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,256,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,128,0.00332622230052948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,256,0.022779555784331426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,128,0.01183733344078064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,64,0.0032542221662071017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,128,0.020633776982625324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,4096,32,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,64,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,65536,0.1307271056705051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,4096,32,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,65536,0.17219200399186876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,51200,0.09613244401084052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,65536,0.23983200391133627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,51200,0.13957866032918295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,16384,0.034860445393456355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,4096,768,0.024122666981485155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,16384,0.050738665792677135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,51200,0.19152000215318468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,12288,0.02769600020514594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,16384,0.07695200045903523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,12288,0.06244888570573595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,12288,0.03833955526351929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,10240,0.03368799885114034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,10240,0.056299553977118603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,8192,0.019630221856964957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,8192,0.04951555530230204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,8192,0.029614223374260798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,7168,0.017514665921529133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,7168,0.04668622215588888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,7168,0.027445332871543035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,6144,0.01572711103492313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,6144,0.04161511196030511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,10240,0.024638222323523626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,6144,0.023855111665195886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,5120,0.013841778039932251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,4096,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,5120,0.021771555145581562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,5120,0.03949955436918471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,4096,0.019398222366968792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,3584,0.01089244418674045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,4096,0.03510133425394694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,3072,0.007970666719807519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,3584,0.017975111802419026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,3584,0.032291554742389254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,2560,0.006959111326270633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,3072,0.01647644407219357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,2560,0.01570044457912445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,2560,0.028864890336990356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,2048,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,2048,0.027473777532577515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,2048,0.014398222168286642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,1536,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,1536,0.02552000019285414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,1536,0.013977777626779346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,3072,0.03017066584693061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,1024,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,1024,0.024319110645188227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,768,0.004253333227501975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,768,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,1024,0.004233777936961916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,768,0.023374223046832617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,512,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,512,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,256,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,512,0.02344266739156511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,256,0.012049777640236748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,128,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,256,0.022517333428064983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,128,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,64,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3584,128,0.02038666605949402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,64,0.011737777955002256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3584,32,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,65536,0.09218311309814453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3584,32,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,65536,0.13687911298539904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,65536,0.23882222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,51200,0.10975822475221421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,51200,0.1908257802327474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,16384,0.029514667060640123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,16384,0.04003999961747064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,51200,0.07287555270724826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,12288,0.0203128887547387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,16384,0.07667733563317193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,12288,0.0623262193467882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,12288,0.03139111068513658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,10240,0.01722666621208191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,8192,0.014502222339312235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,8192,0.02472977836926778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,10240,0.05581511391533745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,10240,0.02804888950453864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,7168,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,8192,0.049732443359163075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,6144,0.011737777955002256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,7168,0.023174222972657945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,7168,0.046375112401114575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,5120,0.010484444598356882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,6144,0.020688888099458482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,6144,0.04178399840990702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,4096,0.009321778184837764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,5120,0.018939556346999276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,4096,0.03492533498340183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,4096,0.01718488832314809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,3584,0.00850399997499254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,3584,0.016223111086421542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,3584,0.03221599923239814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,3072,0.008009778128729926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,3072,0.014463111758232117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,3072,0.030823111534118652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,2560,0.007017778025733099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,2560,0.028791109720865887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,2048,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,5120,0.0389066669676039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,2560,0.014179555906189812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,1536,0.005192888693677055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,2048,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,2048,0.027473777532577515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,1024,0.004409777621428172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,1536,0.025655999779701233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,1536,0.013125333521101209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,1024,0.012395555774370829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,768,0.003899555653333664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,1024,0.023789332972632513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,512,0.0035582222044467926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,768,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,768,0.024562666813532513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,256,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,512,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,512,0.023342221975326538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,128,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,256,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,256,0.02167111138502757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,64,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,128,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,3072,32,0.003162666741344664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,64,0.011011555790901184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,3072,32,0.011688888900809817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,65536,0.09440888961156209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,65536,0.13399199644724527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,51200,0.07192799780103895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,65536,0.23990488052368164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,51200,0.10715822378794353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,16384,0.029147555430730183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,3072,128,0.01926577753490872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,16384,0.03831555445988973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,51200,0.19099199771881104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,12288,0.02314222190115187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,16384,0.07622400257322523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,12288,0.030228445927302044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,12288,0.06228622463014391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,10240,0.01631644368171692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,8192,0.013780444032616086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,10240,0.027103111147880554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,10240,0.05549066596561008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,8192,0.024485333098305598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,8192,0.0496711101796892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,7168,0.012715555727481842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,7168,0.023020444644822016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,6144,0.011353777514563667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,7168,0.046458668178982206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,6144,0.0196435550848643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,5120,0.010012444522645738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,6144,0.041767110427220665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,5120,0.017687999539905123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,5120,0.03901066713862949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,4096,0.01661066710948944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,4096,0.03363200028737386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,3584,0.008758222063382467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,4096,0.009384888741705153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,3584,0.0158595558669832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,3072,0.008017777568764156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,3584,0.03228711088498434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,3072,0.02990488873587714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,3072,0.015377778146002026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,2560,0.006983111302057902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,2560,0.028591109646691218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,2560,0.013779555757840475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,2048,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,2048,0.013224889006879596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,2048,0.026798221800062392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,1536,0.005599110904667113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,1536,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,1536,0.026162667406929865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,1024,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,1024,0.024400000770886738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,768,0.00434044458799892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,768,0.012219555675983429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,1024,0.004875555634498596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,768,0.023303111394246418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,512,0.00397866674595409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,512,0.012053333222866058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,512,0.022357333037588332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,256,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,256,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,256,0.022738666998015508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2560,128,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,64,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,64,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2560,32,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,32,0.01163111130396525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,65536,0.06456266509162055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2560,128,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,65536,0.13138043880462646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,51200,0.052348444859186806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,65536,0.23943911658393013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,51200,0.10491110881169637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,16384,0.02276533345381419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,51200,0.1906355619430542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,16384,0.037752889924579196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,12288,0.018048000004556444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,16384,0.07629333602057563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,12288,0.02993511160214742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,12288,0.062160889307657875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,10240,0.01643200053109063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,10240,0.026791110634803772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,10240,0.05548266569773356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,8192,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,8192,0.02407377792729272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,7168,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,7168,0.02258311046494378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,7168,0.046310222811169095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,6144,0.011133333047231039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,6144,0.019476445184813607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,5120,0.009979555176364051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,6144,0.041273779339260526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,8192,0.04957955413394504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,5120,0.01789155602455139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,4096,0.008363555702898238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,4096,0.0331582228342692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,4096,0.016176000237464905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,3584,0.008009778128729926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,3584,0.03123289015558031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,3072,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,3584,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,3072,0.014139556222491793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,3072,0.0295466681321462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,2560,0.006330666856633292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,2560,0.028536889288160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,5120,0.03850311040878296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,2048,0.005444444302055571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,2048,0.026520000563727483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,2048,0.013193777865833707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,1536,0.004909333255555895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,1536,0.025480889611774023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,1024,0.0042204442951414315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,1024,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,1024,0.023370666636361018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,2560,0.013840888937314352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,768,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,512,0.0038568890757030914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,768,0.02310577697224087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,512,0.011789333489206104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,512,0.023802666200531855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,768,0.003953777667548922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,256,0.011398221883508893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,256,0.021334222621387903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,128,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,1536,0.012872888810104795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,256,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,64,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,128,0.010813333094120026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,2048,32,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,2048,128,0.019000889526473153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,65536,0.05088622371355692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,32,0.01108888867828581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,2048,64,0.011367110742463006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,51200,0.04118400149875217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,65536,0.1281377739376492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,65536,0.23912533124287924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,51200,0.10247911347283258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,16384,0.017492444978819955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,51200,0.19014310836791992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,12288,0.015275556180212231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,16384,0.036908444431093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,12288,0.0620408919122484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,10240,0.013958222336239286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,12288,0.029556443293889362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,10240,0.02679022153218587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,8192,0.013215999636385175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,10240,0.05526933405134413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,8192,0.023584889041052923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,16384,0.07573778099483915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,7168,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,8192,0.049217777119742505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,6144,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,7168,0.04593955477078756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,7168,0.02104444470670488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,5120,0.009395555489593083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,6144,0.019181332654423185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,6144,0.040779554181628756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,5120,0.016957332690556843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,4096,0.008651555412345463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,5120,0.03664799862437778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,4096,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,4096,0.03302933441268073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,3584,0.007296000089910295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,3072,0.006940444310506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,3584,0.03163289030392965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,3584,0.01498311095767551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,2560,0.005961777849329843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,3072,0.014146667387750415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,3072,0.02985422147644891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,2048,0.00552977787123786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,2560,0.0284826656182607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,2048,0.013400889105266996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,1536,0.004650666481918759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,2048,0.027034666803148057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,1536,0.012759110993809171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,1024,0.004020444634887907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,1536,0.02477866742346022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,1024,0.012186666329701742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,768,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,1024,0.024488889508777197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,2560,0.01349688900841607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,512,0.003587555554178026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,768,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,768,0.023711999257405598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,256,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,512,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,512,0.02370400064521366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,128,0.002951111023624738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,256,0.021374222305085924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,256,0.0111191107167138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,64,0.002951111023624738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,128,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1536,128,0.018591110905011494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1536,32,0.002943111169669363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,65536,0.03749777873357137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,64,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,65536,0.1257902251349555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,51200,0.030834668212466772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,65536,0.23872621854146323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,51200,0.10001955429712932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,16384,0.01443288889196184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,51200,0.19059199757046172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,16384,0.03658577799797058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,12288,0.01221244451072481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1536,32,0.010827555424637265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,16384,0.07527911000781588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,10240,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,12288,0.029072887367672388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,12288,0.06185866726769341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,8192,0.010606221854686737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,10240,0.025683555338117812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,10240,0.055008000797695585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,7168,0.010192888478438059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,8192,0.02277155551645491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,8192,0.047373334566752114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,6144,0.010000000397364298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,7168,0.021018667353524104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,7168,0.04391377833154467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,5120,0.008730666504965888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,6144,0.03928000066015456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,6144,0.018417777286635507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,5120,0.016943110360039603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,4096,0.007636444436179266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,5120,0.036425779263178505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,4096,0.015592889653311836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,3584,0.006983111302057902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,4096,0.03253422180811564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,3584,0.014873777826627096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,3072,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,3584,0.031202667289310034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,3072,0.014096889230940076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,2560,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,3072,0.029860443539089624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,2560,0.01350400017367469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,2048,0.005209777918126848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,2560,0.028216888507207234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,2048,0.012924444344308643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,1536,0.004650666481918759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,2048,0.026885333988401625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,1536,0.01240711079703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,1024,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,1536,0.02476088868247138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,1024,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,1024,0.024449777272012498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,768,0.003912888881233004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,768,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,512,0.003624000069167879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,512,0.011737777955002256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,256,0.003338666632771492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,512,0.02236000034544203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,256,0.02130577796035343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,256,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,128,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,128,0.01148888882663515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,768,0.022335110439194575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,64,0.0029315555261241067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,1024,128,0.01899377836121453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,1024,32,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,65536,0.028802666399214003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,64,0.0107360002067354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,1024,32,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,51200,0.02425688836309645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,65536,0.12313066588507758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,65536,0.23986132939656576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,16384,0.015575110912322998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,51200,0.0991520020696852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,51200,0.1901608837975396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,16384,0.03567555546760559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,12288,0.013185777597957186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,16384,0.07534311215082805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,12288,0.028203556934992473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,10240,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,12288,0.06185422341028849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,8192,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,10240,0.02457155618402693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,10240,0.0539537767569224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,7168,0.010024888647927178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,8192,0.02168088820245531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,8192,0.04654844601949056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,6144,0.009291555318567488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,7168,0.043135109874937266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,7168,0.020571556356218126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,5120,0.008655110994974772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,6144,0.018222222725550335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,6144,0.03869600097338358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,4096,0.007733333441946242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,5120,0.03639111253950331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,5120,0.01705777810679542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,4096,0.015801777442296345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,4096,0.03234488765398661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,3584,0.015141333142916361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,3072,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,3584,0.031202667289310034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,3072,0.029503110382292006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,2560,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,2560,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,2560,0.028173334068722192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,2048,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,3584,0.006990222467316522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,2048,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,1536,0.004560889055331548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,2048,0.027080888549486797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,1536,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,1024,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,1536,0.024650666448805068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,3072,0.014126222994592456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,768,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,1024,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,1024,0.024496000674035814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,512,0.00360977773865064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,768,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,768,0.022289777795473736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,256,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,512,0.0233608881632487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,256,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,512,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,256,0.022296888960732356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,768,128,0.018909333480728995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,128,0.01111644423670239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,64,0.0026817777090602448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,64,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,32,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,768,32,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,65536,0.023959111836221483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,768,128,0.003141333245568805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,65536,0.12320978111690944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,51200,0.019273777802785236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,65536,0.2396684487660726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,51200,0.19085689385732016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,51200,0.09751200013690525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,16384,0.01236533373594284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,16384,0.07540533277723524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,16384,0.0353724459807078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,12288,0.010633777413103314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,12288,0.06059466467963325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,12288,0.028604444530275133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,10240,0.010386666489972008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,8192,0.009142222503821055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,10240,0.02530488868554433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,10240,0.05277422070503235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,7168,0.008340444829728868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,8192,0.04660888844066196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,7168,0.0203493336836497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,6144,0.007616000043021307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,7168,0.04318133327696058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,6144,0.03909066650602553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,6144,0.01831644442346361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,5120,0.008039111064540016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,5120,0.036044445302751325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,5120,0.016944888565275405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,4096,0.00776622196038564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,4096,0.03228888909022013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,4096,0.015584889385435315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,3584,0.007078222102589077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,8192,0.02193333374129401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,3584,0.03103644318050808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,3072,0.006309333360857434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,3584,0.014881778094503613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,2560,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,3072,0.013850666582584381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,3072,0.029863999949561224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,2560,0.013507555756303998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,2048,0.0052435555391841466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,2560,0.028539554940329656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,1536,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,2048,0.012882666455374824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,2048,0.02752888864941067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,1024,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,1536,0.02469066613250309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,1024,0.024467556013001338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,1024,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,768,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,768,0.023108444280094568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,768,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,512,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,512,0.022053332792388067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,512,0.011855999628702799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,256,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,1536,0.012488000094890594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,128,0.002943111169669363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,256,0.021360889077186584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,256,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,128,0.010743111371994019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,64,0.0029253332565228143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,512,32,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,512,128,0.018983999888102215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,64,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,65536,0.019121777680185106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,512,32,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,51200,0.016158221496476066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,65536,0.1210906638039483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,65536,0.2394639915890164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,16384,0.00963644435008367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,51200,0.09641866551505195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,51200,0.19101066059536406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,12288,0.008055999875068665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,16384,0.034006221426857844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,16384,0.07431555456585355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,12288,0.027186666925748188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,10240,0.009005332986513773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,12288,0.059192001819610596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,8192,0.008017777568764156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,10240,0.02424266603257921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,10240,0.05312888820966085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,7168,0.007976888782448238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,8192,0.0464053319560157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,8192,0.021710221966107685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,6144,0.0074346669846110875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,7168,0.04286844531695048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,7168,0.020430222153663635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,6144,0.01834222177664439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,5120,0.007694222033023834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,6144,0.03881244526969062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,4096,0.007348444726732042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,5120,0.016904888881577384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,4096,0.015186667442321777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,5120,0.036822223001056247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,3584,0.0069875551594628235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,4096,0.032313777340783015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,3584,0.030975110001034204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,3072,0.006680888848172293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,3584,0.01514933341079288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,3072,0.014098667436175875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,2560,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,3072,0.03011111087269253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,2560,0.01351999988158544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,2560,0.028831998507181805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,2048,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,2048,0.013208000196350945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,2048,0.026916444301605225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,1536,0.004401777767472797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,1536,0.012367111113336353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,1536,0.025103999508751765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,1024,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,1024,0.0248026665714052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,768,0.003527111063400904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,768,0.012150222228633033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,1024,0.003916444463862313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,512,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,768,0.024127110838890076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,512,0.01149511088927587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,256,0.0032720000793536506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,512,0.023151111271646287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,256,0.011160888605647616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,256,0.022446221775478784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,128,0.0029893333299292457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,64,0.00263288886182838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,128,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,256,128,0.019997333486874897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,256,32,0.0029475556479560006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,64,0.010811555716726514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,65536,0.014363555444611443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,256,32,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,51200,0.013070222404268054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,65536,0.11941956149207221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,65536,0.23933866288926867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,16384,0.009460444251696268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,51200,0.09395733144548203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,51200,0.1902693377600776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,12288,0.008097777764002482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,16384,0.07350044118033515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,12288,0.02759733299414317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,10240,0.008789333204428354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,12288,0.05958133273654514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,10240,0.05272622240914238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,10240,0.02453511125511593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,8192,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,8192,0.02148888839615716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,8192,0.04675288995107015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,16384,0.03341066506173875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,7168,0.007713778151406183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,6144,0.007363555332024892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,7168,0.04305600126584371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,7168,0.020319110817379422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,5120,0.007263110743628607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,6144,0.018209778600268893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,6144,0.03902577691608005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,4096,0.00675999994079272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,5120,0.01687733332316081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,5120,0.03640800052218967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,3584,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,4096,0.015581332974963717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,4096,0.03299555513593886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,3072,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,3584,0.03136000037193298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,3584,0.014851555228233337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,2560,0.006309333360857434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,3072,0.014151111245155334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,3072,0.0291884442170461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,2048,0.006001777946949005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,2560,0.01351199961370892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,2560,0.028879112667507593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,2048,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,1536,0.004523555437723796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,2048,0.026879999372694228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,1536,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,1024,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,1536,0.025182222326596577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,768,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,1024,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,1024,0.024496888120969135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,512,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,768,0.023064000738991633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,768,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,256,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,512,0.023772444989946153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,512,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,128,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,256,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,256,0.02204177776972453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,64,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,128,0.010856000085671743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,192,128,128,0.01996977792845832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,128,32,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,64,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,65536,0.013547555440002017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,51200,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,128,32,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,65536,0.11861777305603027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,16384,0.009023110899660323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,51200,0.09452266825570001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,16384,0.033618665403789945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,12288,0.007900444169839224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,10240,0.007357333269384172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,12288,0.027493332823117573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,8192,0.006679111056857639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,10240,0.024433778391944036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,7168,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,8192,0.02163022259871165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,6144,0.006337777607970768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,7168,0.020299555526839364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,6144,0.01834133267402649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,5120,0.006264888991912206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,4096,0.00563822231358952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,5120,0.01689511040846507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,3584,0.0063075555695427795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,4096,0.015240000353919135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,3584,0.014897776974572076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,3072,0.014164444473054675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,2560,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,2560,0.013435555828942193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,2048,0.005979555762476391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,2048,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,1536,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,1536,0.012463111016485425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,3072,0.005980444451173146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,1024,0.003823111040724648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,768,0.003549333247873518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,1024,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,512,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,512,0.011517333487669626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,256,0.002945777856641345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,256,0.011137777732478248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,128,0.002663111107216941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,128,0.010710222025712332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,64,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,768,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,64,32,0.002507555608948072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,64,0.010832000109884473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,65536,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,64,32,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,51200,0.011720889144473605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,65536,0.1195146640141805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,16384,0.008513777620262569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,51200,0.09238311317231919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,12288,0.007703999678293864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,16384,0.03334488802485996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,12288,0.027447111076778833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,10240,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,8192,0.006675555474228329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,7168,0.006545777950021956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,10240,0.024360888534122046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,6144,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,8192,0.02141511109140184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,6144,0.01828977796766493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,5120,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,4096,0.005767999837795894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,5120,0.016799110505315993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,3584,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,4096,0.01515111161602868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,3072,0.005979555762476391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,3584,0.014787554740905762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,2560,0.006272000157170826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,3072,0.013533333109484779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,2048,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,2560,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,1536,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,2048,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,1024,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,1536,0.012379555238617791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,768,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,7168,0.02028622229894002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,1024,0.011708444191349877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,768,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,256,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,512,0.011160888605647616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,128,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,256,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,64,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,128,0.010397333237859938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,32,0.002530666689078013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,64,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,192,32,32,0.010417777631017896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,192,32,512,0.0032426667296224167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,16384,0.39412087864345974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,51200,0.8541075388590494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,51200,1.1434692806667752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,12288,0.285974211162991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,16384,0.27704177962409127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,12288,0.2078240050209893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,10240,0.2368097835116916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,8192,0.19408533308241102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,10240,0.17535732852088082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,8192,0.14065066973368326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,7168,0.16614666250016955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,6144,0.14620533254411486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,7168,0.12422577540079753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,6144,0.10874311129252116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,5120,0.12197689215342204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,4096,0.09781333472993638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,5120,0.09253866804970635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,4096,0.07704977856742011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,3584,0.08619022369384766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,3072,0.07506133450402154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,2560,0.06399910979800753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,3584,0.06923644410239325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,3072,0.062041779359181724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,2048,0.0524453322092692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,2560,0.053872889942593045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,1536,0.04014577799373203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,2048,0.04672000143263075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,1024,0.02866577770974901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,1536,0.03898222247759501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,768,0.023256889647907678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,512,0.016394666499561734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,1024,0.03214577833811442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,256,0.010451555252075195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,768,0.03128800127241347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,128,0.009519110951158736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,256,0.02164977788925171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,512,0.02438933319515652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,64,0.008998221821255153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,65536,32,0.009717333647939894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,128,0.02010844482315911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,64,0.019663110375404358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,65536,32,0.02176800039079454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,65536,1.0967831081814237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,65536,0.8639075491163465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,51200,0.8637324439154731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,16384,0.2945946587456597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,12288,0.2170648839738634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,16384,0.2207875516679552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,51200,0.6778808699713813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,10240,0.20143377780914307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,12288,0.16731733745998809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,8192,0.16603555944230822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,7168,0.14173689153459337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,8192,0.11433245076073541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,7168,0.10088088777330186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,6144,0.12438044283125137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,5120,0.10428800185521443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,6144,0.08795644177330865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,10240,0.14122222529517278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,4096,0.08534044689602321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,5120,0.07559466361999512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,3584,0.0758719974093967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,4096,0.06344711118274264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,3584,0.057189332114325635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,3072,0.05090044604407417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,2560,0.05680355760786268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,2048,0.04715199934111702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,2560,0.04504444532924228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,1536,0.03194844391610887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,2048,0.03949866692225138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,1024,0.023003555006451074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,1536,0.03357244531313578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,768,0.018087999688254464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,1024,0.02682488825586107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,768,0.026397332549095154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,3072,0.06619289186265734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,512,0.013251555462678274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,256,0.009038222332795462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,128,0.008601778083377415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,512,0.021673777037196692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,64,0.007339555356237624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,256,0.019448000523779128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,51200,32,0.007675555845101674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,64,0.01819022165404426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,128,0.018333334061834548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,51200,32,0.01905066602759891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,65536,0.3655573262108697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,51200,0.2854915459950765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,65536,0.35065333048502606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,16384,0.09921511014302571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,51200,0.2754257784949409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,12288,0.07318311267428927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,16384,0.095652441183726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,12288,0.07335378064049615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,10240,0.06295200188954671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,8192,0.052242666482925415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,7168,0.04490400022930569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,10240,0.06469421916537814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,8192,0.05387644304169548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,6144,0.04047111007902358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,7168,0.04617599977387322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,5120,0.034120000070995755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,4096,0.0277493331167433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,5120,0.03639111253950331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,6144,0.04124266571468777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,3584,0.025063999825053748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,3072,0.022171555293930903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,4096,0.03188088867399428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,3072,0.025425778494940862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,3584,0.02920000089539422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,2560,0.019173334042231243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,2048,0.0159608887301551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,1536,0.012860444684823355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,2560,0.023649776975313824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,1024,0.009878222313192155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,2048,0.021334222621387903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,1536,0.019179556104871962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,768,0.00870222184393141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,768,0.016951110627916124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,512,0.006816888848940532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,512,0.014584889014561972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,256,0.004628444297446145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,1024,0.016876444220542908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,128,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,256,0.013841778039932251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,64,0.003940444439649582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,128,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,16384,32,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,64,0.012829333543777466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,16384,32,0.013468444347381592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,65536,0.2720453209347195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,65536,0.26575732231140137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,51200,0.213137772348192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,16384,0.07667555411656697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,51200,0.21034044689602324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,12288,0.055889778667026095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,16384,0.0755315555466546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,12288,0.05834577480951945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,10240,0.04793777730729845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,8192,0.04051466782887777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,7168,0.0347306662135654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,8192,0.0439297788672977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,6144,0.030813333061006334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,7168,0.037831111086739436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,5120,0.02668711046377818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,6144,0.0336408880021837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,5120,0.030402666992611353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,4096,0.021914665897687275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,3584,0.019708444674809773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,4096,0.02585244509908888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,3072,0.01696266730626424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,3584,0.023761777414215937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,10240,0.05140711201561821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,2560,0.014833778142929077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,3072,0.022220444348123338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,2048,0.012977778083748288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,1536,0.010612444745169746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,2560,0.020412445068359375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,1024,0.008664000365469191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,2048,0.01864888932969835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,768,0.007435555259386699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,1024,0.014919999572965833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,1536,0.01724799970785777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,512,0.005962666538026597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,768,0.015400000744395785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,256,0.004937777916590373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,128,0.004315555509593752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,256,0.013215999636385175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,512,0.013878222141000958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,64,0.003937777959638172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,128,0.012470222181744046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,12288,32,0.0042568888101312844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,32,0.013955555028385587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,12288,64,0.01218844453493754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,65536,0.2681395477718777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,51200,0.23240534464518228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,65536,0.2556684547000461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,51200,0.2038604418436686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,16384,0.07878133323457506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,12288,0.05471644467777676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,10240,0.04519733455446032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,16384,0.07309777869118585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,8192,0.04375466704368591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,12288,0.05687644746568468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,10240,0.050458666351106435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,7168,0.03430666526158651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,8192,0.042464001311196216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,6144,0.032280001375410296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,5120,0.026781333817376032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,6144,0.03202844328350491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,4096,0.022094221578703985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,5120,0.02801333202256097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,3584,0.020978666014141504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,4096,0.02458222210407257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,3072,0.018637332651350234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,3584,0.02342133389578925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,3072,0.02156266735659705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,2560,0.017643555998802185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,2048,0.013751111096805997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,2560,0.01995022263791826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,1536,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,7168,0.03635111120012071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,2048,0.01830844415558709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,1536,0.016757332616382174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,768,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,1024,0.014503111441930136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,768,0.01460533340771993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,512,0.005647999958859549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,256,0.004601777841647466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,512,0.01350844403107961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,128,0.004224888980388641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,256,0.012816888590653738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,128,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,64,0.004288889053795072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,32,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,64,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,10240,32,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,10240,1024,0.00829777783817715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,65536,0.2115093337164985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,65536,0.21466310818990073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,51200,0.16592355569203696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,16384,0.058360000451405845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,51200,0.17142132918039957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,12288,0.044201778040991895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,16384,0.06287378072738647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,12288,0.049831112225850425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,10240,0.03812711106406318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,8192,0.03171111146608988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,10240,0.04352800051371256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,7168,0.027698665857315063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,8192,0.03575999869240655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,6144,0.024706666668256123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,7168,0.03158044483926561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,5120,0.02145244512293074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,5120,0.02477866742346022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,6144,0.028667555914984807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,4096,0.01759111053413815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,3584,0.015565334094895257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,4096,0.022397332721286353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,3584,0.021036444438828364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,3072,0.013800000150998434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,2560,0.012229333321253458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,3072,0.01992266707950168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,2048,0.01073333372672399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,2560,0.018650665879249573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,1536,0.008747555315494537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,2048,0.017216889394654166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,1024,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,1536,0.01550222271018558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,1024,0.013836444252067141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,768,0.005249777601824866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,512,0.004592000196377437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,768,0.014170666535695394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,256,0.0041644444896115195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,512,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,128,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,256,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,64,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,128,0.011864888999197217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,64,0.012435555458068848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,8192,32,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,8192,32,0.013456888496875763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,65536,0.20026755332946777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,51200,0.15732444657219782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,65536,0.21011555194854736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,16384,0.05561777618196276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,16384,0.061623109711541064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,51200,0.1684044467078315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,12288,0.04246133234765795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,10240,0.03677422139379714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,8192,0.030833777454164293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,12288,0.048787554105122886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,7168,0.02640266716480255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,10240,0.04264266623391045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,6144,0.02346755564212799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,7168,0.03066666589842902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,8192,0.03451111250453525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,5120,0.019667555888493855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,4096,0.016354666815863717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,6144,0.027407111393080816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,3584,0.01480888823668162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,4096,0.02236088944805993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,5120,0.024774221910370722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,3072,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,2560,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,3584,0.020974222156736586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,2048,0.010370666782061258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,3072,0.019706666469573975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,2560,0.018234666850831773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,1536,0.008593777815500895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,1024,0.006696888970004187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,2048,0.016876444220542908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,1536,0.0151573336786694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,768,0.0052897776994440295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,512,0.004572444491916233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,1024,0.01383733335468504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,768,0.014112888111008538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,256,0.003921777837806278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,128,0.003903111235962974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,256,0.01239377756913503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,512,0.01315822203954061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,64,0.0035128887328836652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,7168,32,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,128,0.01181333346499337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,64,0.012163555456532372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,7168,32,0.013097777962684631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,65536,0.17288977570003936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,51200,0.13598755995432535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,16384,0.04570577873124016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,65536,0.17447022596995035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,12288,0.03484888871510824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,51200,0.1386257807413737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,16384,0.05232977867126465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,10240,0.030223998758527968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,8192,0.024902222885025874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,12288,0.0418835547235277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,10240,0.035713778601752386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,7168,0.021716444028748408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,8192,0.02927466564708286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,7168,0.02593066626124912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,6144,0.019359999232821994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,5120,0.01668177710639106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,6144,0.023762666516833838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,4096,0.013549333645237817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,5120,0.021762667430771723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,4096,0.019985778464211356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,3584,0.012284444438086616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,3072,0.011169777976142036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,3584,0.018927999668651156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,2560,0.009768889182143742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,3072,0.01791111131509145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,2048,0.008758222063382467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,2560,0.016914667354689706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,1536,0.007436444362004598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,2048,0.014855111638704935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,1536,0.013896889156765409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,1024,0.00600355573826366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,768,0.005034666508436203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,1024,0.01314133322901196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,512,0.0043333334227403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,768,0.01350844403107961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,256,0.003995555556482739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,512,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,128,0.0035582222044467926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,128,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,256,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,64,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,6144,32,0.0033297776761982176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,64,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,6144,32,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,65536,0.14549333519405788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,65536,0.17222400506337485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,51200,0.11496178309122722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,16384,0.04247911108864678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,51200,0.13701600498623318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,12288,0.032644444041781955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,16384,0.050711999336878456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,10240,0.028157333532969158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,12288,0.03904622130923801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,8192,0.023194667365815904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,10240,0.03358577688535055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,7168,0.0203804439968533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,8192,0.028591109646691218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,6144,0.017702221870422363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,6144,0.023618666662110224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,7168,0.025324443976084392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,5120,0.01536177761024899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,4096,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,5120,0.021504888931910198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,3584,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,3584,0.018798222144444782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,3072,0.010727110836240979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,3072,0.01735911104414198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,2560,0.009677333136399588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,4096,0.01961599952644772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,2560,0.015968888998031616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,2048,0.008349333372380998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,1536,0.007056000332037608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,2048,0.014899555179807874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,1536,0.014195554786258273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,1024,0.005530666559934616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,768,0.00463733325401942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,512,0.004271111140648524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,1024,0.01311555587583118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,256,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,768,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,128,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,512,0.012532444463835822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,256,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,64,0.003481777591837777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,128,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,5120,32,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,64,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,65536,0.10973778035905625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,5120,32,0.011952000359694162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,51200,0.0865280032157898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,65536,0.16418400075700548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,16384,0.035347554418775774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,12288,0.028600000672870215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,51200,0.13195733229319254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,10240,0.02307111190425025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,16384,0.04785155587726169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,8192,0.01870044403605991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,10240,0.032256000571780734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,7168,0.01664977769056956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,8192,0.028359999259312946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,6144,0.014608888162506951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,7168,0.024855110380384658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,5120,0.013073777986897362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,6144,0.023016000787417095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,4096,0.009741333623727163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,5120,0.021291555629836187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,4096,0.0192284451590644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,3584,0.017988445030318368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,3072,0.008551999926567078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,12288,0.03727022144529555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,3072,0.016336888074874878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,2560,0.00777688870827357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,2560,0.01548266741964552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,2048,0.006647999915811751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,2048,0.014527999692493014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,3584,0.009293333523803288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,1536,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,1024,0.004585777719815572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,1536,0.013678222066826291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,768,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,1024,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,512,0.0038737778862317405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,768,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,512,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,256,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,128,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,256,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,64,0.0033493333806594214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,64,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,4096,32,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,32,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,65536,0.10381333033243816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,65536,0.16284533341725668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,51200,0.08171021938323975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,4096,128,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,16384,0.03363022208213806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,51200,0.13050488630930582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,12288,0.0269795556863149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,16384,0.047129776742723256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,10240,0.025088888075616624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,12288,0.035967999034457736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,8192,0.02023555503951179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,10240,0.0321688883834415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,7168,0.01960177719593048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,8192,0.02808622188038296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,6144,0.01570044457912445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,7168,0.02511111067401038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,5120,0.01442577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,6144,0.022978666755888198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,4096,0.012058667010731168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,5120,0.020971554848882887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,3584,0.01165688865714603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,4096,0.018872888551818002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,3072,0.009751111268997192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,3584,0.01679733395576477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,2560,0.009287111461162567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,3072,0.016494222813182406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,2048,0.007839110990365347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,2560,0.015187554889255099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,1536,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,2048,0.014447111222479077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,1024,0.005568000177542369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,1536,0.013759111364682516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,768,0.004853333450025982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,1024,0.013040888640615674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,512,0.0042462220622433555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,768,0.01310400002532535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,256,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,512,0.01240355521440506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,128,0.003375111147761345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,256,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,64,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,128,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3584,32,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,64,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3584,32,0.011798222031858234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,65536,0.1257893376880222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,51200,0.06462399827109443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,51200,0.10125066836675008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,16384,0.02737688852681054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,16384,0.03717333409521315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,12288,0.021846221552954778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,65536,0.0799600018395318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,10240,0.018586667047606576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,12288,0.029167999823888142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,8192,0.014863999353514777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,7168,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,10240,0.02641777859793769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,6144,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,8192,0.02404800057411194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,7168,0.020976889464590285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,5120,0.010776888993051318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,4096,0.009567110902733272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,6144,0.0195777780479855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,5120,0.017683555682500202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,3584,0.008985777695973715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,3072,0.008024888734022776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,4096,0.016179554992251925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,2560,0.007320000065697565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,3584,0.015156444576051501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,2048,0.006205333189831839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,3072,0.01442311041884952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,1536,0.0053324442770746015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,2560,0.014163555370436775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,1024,0.004616888860861461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,2048,0.013438222308953604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,768,0.0038613333470291565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,768,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,1024,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,512,0.0035386664999855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,512,0.012732444538010491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,256,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,256,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,128,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,1536,0.013067555924256643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,64,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,64,0.010724444356229572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,3072,32,0.003148444410827425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,32,0.011305777562989129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,65536,0.073725336127811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,65536,0.1225315597322252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,51200,0.059848891364203564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,51200,0.09856622086630927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,3072,128,0.011653333074516721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,16384,0.02588533361752828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,12288,0.02046133412255181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,16384,0.0365866654449039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,12288,0.028775999943415325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,10240,0.020024889045291476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,8192,0.01475644442770216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,10240,0.025969778498013813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,7168,0.013208000196350945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,7168,0.02064622276359134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,6144,0.012055111428101858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,8192,0.023436443673239812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,5120,0.010381333529949188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,6144,0.01885688967174954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,4096,0.009352888498041365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,5120,0.017398221625222098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,3584,0.009043555292818282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,4096,0.015831111205948722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,3072,0.008427555362383524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,3584,0.014861333701345654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,3072,0.014160888062583076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,2560,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,2048,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,2048,0.013071999781661563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,1536,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,1536,0.012816000315878125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,1024,0.004933333231343163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,2560,0.006994666324721442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,1024,0.012228444218635559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,768,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,512,0.003863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,768,0.012461333639091916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,256,0.0036106668412685394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,512,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,128,0.0034097778714365433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,256,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,64,0.003134222287270758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,64,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2560,32,0.003030222323205736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,128,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2560,32,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,65536,0.06143822272618612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,51200,0.048475556903415255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,65536,0.1189182202021281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,51200,0.09614133172565037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,12288,0.01773955590195126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,16384,0.0347377790345086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,10240,0.016176000237464905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,12288,0.028399109840393066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,8192,0.014388443695174323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,10240,0.025528000460730657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,7168,0.012545777691735161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,8192,0.023016888234350417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,7168,0.01997599999109904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,6144,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,5120,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,6144,0.018262222409248352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,4096,0.008908444808589088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,5120,0.017250667015711468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,16384,0.02257333364751604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,3584,0.007986666427718269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,4096,0.015488889482286243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,3072,0.0075839997993575195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,2560,0.0063182223174307085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,3584,0.014871110518773397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,2048,0.0058142224119769196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,2560,0.013832888669437833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,1536,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,2048,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,1024,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,1536,0.012924444344308643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,768,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,1024,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,768,0.012574222352769641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,512,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,256,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,512,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,3072,0.013985777894655863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,128,0.003152888889114062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,64,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,256,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,2048,32,0.002944888960984018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,128,0.011040888726711273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,65536,0.045824001232783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,64,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,51200,0.037767112255096436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,2048,32,0.011109333071443768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,65536,0.11626399887932672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,16384,0.016750221451123554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,51200,0.09377421935399373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,12288,0.01402311192618476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,16384,0.034338666333092585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,10240,0.013904889424641928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,12288,0.028041776683595445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,10240,0.025361778007613286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,8192,0.012249777714411417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,7168,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,8192,0.021770666042963665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,6144,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,7168,0.019243554936514962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,5120,0.009524444739023844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,6144,0.017717333303557504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,4096,0.008002666963471307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,5120,0.016555555992656283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,3584,0.007618666523032718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,4096,0.015519110692871941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,3072,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,3584,0.014785778191354541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,3072,0.013820444544156393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,2560,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,2048,0.0052826665341854095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,2560,0.013474666410022311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,1536,0.004876444323195351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,2048,0.013439999686347114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,1024,0.004206222378545337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,1536,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,768,0.003899555653333664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,1024,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,512,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,768,0.012113778127564324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,256,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,128,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,512,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,64,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,128,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,256,0.011514667007658215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1536,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,65536,0.03348533312479655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,32,0.011129777464601727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1536,64,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,51200,0.027875555886162653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,65536,0.11380355887942845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,16384,0.013742222554153867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,12288,0.012359999948077731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,51200,0.09089866611692642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,16384,0.03386044502258301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,12288,0.027477333943049114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,8192,0.01036266651418474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,10240,0.02438755498992072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,7168,0.009692444569534725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,8192,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,7168,0.018922666708628338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,6144,0.010100444157918295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,5120,0.008665777742862701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,6144,0.017587555779351127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,4096,0.00757244477669398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,5120,0.01648266613483429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,3584,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,4096,0.01516533394654592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,3072,0.006274666637182236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,3584,0.014678221609857349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,2560,0.005622222191757626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,10240,0.011370666325092316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,3072,0.013776000175211163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,2048,0.005399999933110342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,2560,0.013555555707878537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,1536,0.00518844442235099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,2048,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,1024,0.004332444320122401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,1536,0.012709333664841123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,768,0.004009777886999978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,1024,0.01204355557759603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,512,0.0036275556517971885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,768,0.012089778151777057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,256,0.003338666632771492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,512,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,128,0.003664888855483797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,256,0.011696889168686338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,64,0.003310222178697586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,128,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,1024,32,0.0030337776988744736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,64,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,65536,0.028992888000276353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,1024,32,0.011016888750924004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,51200,0.023757333556811016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,16384,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,65536,0.11218844519721137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,12288,0.013199110825856527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,51200,0.0898853341738383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,10240,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,16384,0.03376266691419814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,8192,0.011022222538789114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,12288,0.026240888569090102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,7168,0.009687999884287516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,10240,0.023543111152119104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,6144,0.009295110901196798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,8192,0.02120977805720435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,7168,0.01871288816134135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,5120,0.00796977761718962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,5120,0.016532444291644625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,4096,0.007278222176763747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,4096,0.015196444259749519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,3584,0.0069511110583941145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,3584,0.01479466590616438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,3072,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,3072,0.01385600037044949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,2560,0.005618666609128316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,2560,0.01353422221210268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,2048,0.0052471111218134565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,2048,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,1536,0.004569777597983678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,1536,0.012803555362754397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,1024,0.00423911131090588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,1024,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,768,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,6144,0.01757955551147461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,512,0.0034186666210492453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,768,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,512,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,128,0.00295733329322603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,256,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,64,0.0027280000762807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,128,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,768,32,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,64,0.011036444041464063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,65536,0.023717333873112995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,768,32,0.011026666396194033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,51200,0.01963644391960568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,65536,0.11111288600497776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,16384,0.012318222059143914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,51200,0.08826311429341634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,12288,0.010668444136778513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,16384,0.03243466549449497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,10240,0.009682666924264696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,12288,0.02612888813018799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,8192,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,10240,0.0233288887474272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,7168,0.007998222278224098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,8192,0.020621332857343886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,6144,0.00868000007337994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,7168,0.018904000520706177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,5120,0.00828888929552502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,6144,0.017536888519922893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,4096,0.007473777565691207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,5120,0.01645599967903561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,3584,0.0069422221018208405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,4096,0.015132443772421943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,3072,0.006599111275540457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,3584,0.014476444986131458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,2560,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,3072,0.013807111316257052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,2048,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,2560,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,1536,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,2048,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,1024,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,1536,0.012695111334323883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,768,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,1024,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,512,0.0034977777136696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,768,0.01239911135700014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,256,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,512,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,128,0.00300799993177255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,256,0.011720889144473605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,64,0.002559111143151919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,128,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,512,32,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,64,0.011042666931947073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,65536,0.018908444378111098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,512,32,0.010739555789364709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,51200,0.016224000189039443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,65536,0.1091724435488383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,16384,0.011619555453459421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,12288,0.009684444301658207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,51200,0.08583466874228583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,10240,0.008966222405433655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,16384,0.031529777579837374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,8192,0.007990222010347579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,12288,0.025705777936511572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,10240,0.023033777872721355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,7168,0.007673777639865875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,8192,0.020785777105225455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,6144,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,5120,0.007689778175618913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,7168,0.018594667315483093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,4096,0.007317333585686154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,6144,0.01755288905567593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,3584,0.00666844430896971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,5120,0.016252444850073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,4096,0.014913777510325113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,3072,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,3584,0.014523555835088095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,2560,0.005594666633341048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,2048,0.0052391112678580815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,3072,0.013526221944226159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,2560,0.013449778159459433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,1536,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,2048,0.012849777936935425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,1024,0.003928889003064898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,1536,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,768,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,1024,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,768,0.012155555188655853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,512,0.003403555601835251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,256,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,512,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,128,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,256,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,64,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,128,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,256,32,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,64,0.010454222559928894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,256,32,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,65536,0.01422755585776435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,51200,0.01556711064444648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,65536,0.10751999749077691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,16384,0.011137777732478248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,51200,0.08526221911112468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,16384,0.03127911024623447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,12288,0.00958933350112703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,10240,0.008742222355471717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,12288,0.025447999437650044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,8192,0.008070222205585903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,10240,0.02304977840847439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,7168,0.008665777742862701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,8192,0.020720889170964558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,6144,0.008041777544551426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,7168,0.018580444984965853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,6144,0.017245334055688646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,5120,0.008341333104504479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,4096,0.007377777662542131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,5120,0.016196444630622864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,3584,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,4096,0.014930667148696052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,3072,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,3072,0.01367911116944419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,2560,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,2560,0.013221333424250284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,2048,0.005151111218664381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,3584,0.014533332652515836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,1536,0.004560889055331548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,2048,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,1024,0.003914666672547658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,1536,0.012463111016485425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,768,0.003588444242874781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,1024,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,768,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,512,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,256,0.003088888815707631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,512,0.012088889049159156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,128,0.00295733329322603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,256,0.011517333487669626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,64,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,128,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,128,32,0.002565333412753211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,32,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,128,64,0.0105608892109659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,65536,0.014087110757827759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,51200,0.01346666696998808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,51200,0.08482933044433594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,65536,0.10792444149653117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,16384,0.009011555876996782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,12288,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,16384,0.03126577867401971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,10240,0.008314666648705801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,10240,0.023052444060643513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,8192,0.007711999946170383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,8192,0.020580444071027968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,7168,0.008728000024954477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,12288,0.025451555848121643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,6144,0.00831822223133511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,7168,0.018636445204416912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,5120,0.00737955586777793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,6144,0.017304889029926725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,4096,0.007015110717879401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,5120,0.016235555211702984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,3584,0.00646311127477222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,4096,0.0151582227812873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,3072,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,3584,0.014527110589875115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,3072,0.013765333427323235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,2560,0.005296000176005893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,2048,0.00499288903342353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,2560,0.013448889056841532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,1536,0.00452977791428566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,2048,0.012792000340090858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,1536,0.012396444049146442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,1024,0.0036106668412685394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,768,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,1024,0.012072000238630505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,512,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,768,0.011904000408119626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,256,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,512,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,128,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,256,0.011543110840850405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,128,0.010805333654085795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,64,0.0029120000286234748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,64,0.010501333408885531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,64,32,0.0026168889469570587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,64,32,0.010861333045694562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,65536,0.013908444179428948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,65536,0.10673511028289795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,51200,0.013222222526868185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,51200,0.08425066868464152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,16384,0.00904888908068339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,16384,0.031319111585617065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,12288,0.009045333498054082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,12288,0.02548177705870734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,10240,0.008723555339707268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,10240,0.02312177750799391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,8192,0.007714666426181793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,8192,0.020482665962643094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,7168,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,6144,0.008230222596062554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,7168,0.01863466699918111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,5120,0.00739733295308219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,6144,0.017251556118329365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,4096,0.007128000259399414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,5120,0.016202666693263583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,3584,0.006680888848172293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,4096,0.014904000692897372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,3072,0.0063075555695427795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,3072,0.01349422252840466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,3584,0.014562666416168213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,2560,0.01316177762216992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,2048,0.004925333377387789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,2048,0.012826667063766055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,1536,0.00424888895617591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,2560,0.005630222045713001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,1536,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,1024,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,768,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,1024,0.01183111137813992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,768,0.011786667009194693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,512,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,256,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,512,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,128,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,256,0.011448889142937131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,64,0.0025724443710512584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,128,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,161,32,32,0.002559111143151919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,64,0.010454222559928894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,161,32,32,0.010504888991514841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,51200,0.8521208763122559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,51200,1.038823127746582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,16384,0.3616320027245416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,16384,0.27351734373304576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,12288,0.261735995610555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,12288,0.20710221926371256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,16384,0.27710045708550346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,10240,0.23566044701470268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,12288,0.2153057787153456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,10240,0.17500178019205728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,8192,0.19413067234887016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,8192,0.1410542196697659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,10240,0.18487021658155653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,7168,0.16573777463701037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,7168,0.12461245059967041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,8192,0.15338488419850668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,6144,0.14559111330244276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,6144,0.10845066441429986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,5120,0.12175022231207953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,51200,0.8094879786173502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,7168,0.13596444659762913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,5120,0.09245066510306464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,4096,0.097944888803694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,4096,0.07668888568878174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,6144,0.12106045087178548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,3584,0.08623466889063518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,5120,0.10572533475028144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,4096,0.0908764468299018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,3072,0.0746648907661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,3584,0.06885244448979695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,3072,0.06167733007007175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,2560,0.06383466720581055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,3584,0.08235022094514635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,2048,0.052173333035575024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,2560,0.053878221246931285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,3072,0.07640444570117526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,1536,0.04021777709325155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,2048,0.04639733168813917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,2560,0.06762666834725274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,1024,0.028507557180192735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,2048,0.06077600187725491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,1536,0.03896977835231357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,1536,0.05348799957169426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,768,0.022828444838523865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,1024,0.03183199961980184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,1024,0.046705779102113515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,512,0.016252444850073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,768,0.02702577743265364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,256,0.010279110736317104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,768,0.04432000054253472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,512,0.02366133365366194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,512,0.038225776619381376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,256,0.021314667330847845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,128,0.008727110922336578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,256,0.03438222077157762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,64,0.007797333101431529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,65536,128,0.027560000618298847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,65536,32,0.00832266691658232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,64,0.01966399947802226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,32,0.02072000006834666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,65536,128,0.019951999187469482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,65536,1.0915706422593858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,65536,0.861769782172309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,51200,0.860496891869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,51200,0.6806097560458713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,16384,0.2914888858795166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,16384,0.2207697762383355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,16384,0.24296532736884224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,12288,0.2170097827911377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,12288,0.1667893330256144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,51200,0.7068524360656738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,12288,0.1887742280960083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,65536,0.8948595258924695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,10240,0.17852355374230278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,8192,0.15037688944074842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,10240,0.14120977454715306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,8192,0.11439377731747097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,7168,0.12599378161960179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,10240,0.16112800439198813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,7168,0.10059555371602376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,6144,0.11175910631815593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,8192,0.13542044162750244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,6144,0.08773422241210938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,5120,0.09248799747890896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,7168,0.12068177594078912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,5120,0.07569511069191827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,6144,0.1075742244720459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,5120,0.09419200155470107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,4096,0.06336177719963922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,3584,0.06711733341217041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,4096,0.08127199941211276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,3584,0.056848002804650195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,3072,0.05841955873701307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,3584,0.0743502246008979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,4096,0.076273779074351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,2560,0.050429334243138633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,3072,0.051148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,2560,0.045328888628217906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,3072,0.06890400250752766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,2048,0.042191998826132886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,2560,0.06173066960440742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,2048,0.039523555172814265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,1536,0.03205777870284186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,2048,0.05552088883188036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,1024,0.023342221975326538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,1536,0.033372445238961115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,1536,0.04909155435032315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,1024,0.0273626661962933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,768,0.01825066738658481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,1024,0.04233066572083367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,512,0.01330044451687071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,768,0.023715555667877197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,768,0.039417776796552874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,256,0.00941955546538035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,512,0.03566666775279575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,512,0.020786666207843356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,128,0.006960888703664144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,256,0.03215822246339586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,256,0.019167999426523846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,128,0.01790488925245073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,51200,128,0.025711110896534387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,32,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,64,0.01752088963985443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,51200,32,0.019253333409627277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,65536,0.36832888921101886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,65536,0.3495466709136963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,51200,0.2895653247833252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,51200,64,0.005655111124118169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,51200,0.2760808997684055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,16384,0.10204178094863892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,16384,0.09554133150312637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,65536,0.3708008925120036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,12288,0.07480977641211616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,16384,0.11087022225062053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,51200,0.2958204481336805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,10240,0.0633573333422343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,12288,0.0734506646792094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,12288,0.08887822098202175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,10240,0.06449777550167508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,8192,0.05293955405553182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,10240,0.0787573324309455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,7168,0.04538044333457947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,8192,0.05391644438107809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,8192,0.06873866584565905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,6144,0.040411555104785495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,7168,0.04656444324387444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,7168,0.06202666627036201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,6144,0.04155822263823615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,5120,0.03462488783730401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,6144,0.05733866824044121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,5120,0.03674844569630093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,4096,0.028460443019866943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,5120,0.051867554585138954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,4096,0.03162044286727905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,3584,0.02542488939232296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,4096,0.046668443414900035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,3584,0.029136889510684546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,3072,0.022460444105996027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,3584,0.04381155636575487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,3072,0.02664711078008016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,2560,0.019520888725916546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,3072,0.04098755452367995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,2560,0.023386667172114056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,2048,0.01631733278433482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,2560,0.038176000118255615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,2048,0.021363554729355708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,1536,0.012925333446926542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,2048,0.035960889524883695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,1536,0.019313777486483257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,1024,0.010225777824719747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,1536,0.033259554041756526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,1024,0.017252445220947266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,768,0.00870577742656072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,1024,0.030544890297783747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,768,0.01590044465329912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,512,0.006703111032644908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,768,0.02953333324856228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,512,0.014635556273990206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,256,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,512,0.026128000683254663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,256,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,128,0.004003555410438114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,256,0.023764444722069636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,128,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,64,0.0038488888078265717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,16384,128,0.020642666353119742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,16384,32,0.003976888954639435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,64,0.013754666679435305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,16384,32,0.01348355578051673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,65536,0.2714284525977241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,65536,0.26480266782972545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,51200,0.21159999900394014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,51200,0.21085333824157715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,65536,0.3691573407914903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,16384,0.07657066980997722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,16384,0.07545333438449435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,16384,0.11053511169221665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,51200,0.2937671078575982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,12288,0.05522844526502821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,12288,0.05858844518661499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,10240,0.047876444127824574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,12288,0.08846133285098606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,8192,0.04020533296797011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,10240,0.052051554123560585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,10240,0.07868978050020006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,8192,0.043818665875328906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,7168,0.03436000148455302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,8192,0.06835022237565783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,7168,0.03784888982772827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,6144,0.030934221214718286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,7168,0.06227022409439087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,5120,0.026455110973782007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,6144,0.0337759984864129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,6144,0.056886222627427846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,4096,0.022074666288163926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,5120,0.030213332838482324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,5120,0.05123200019200643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,3584,0.019507555498017203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,4096,0.046275556087493896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,3584,0.02368444369898902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,4096,0.025669333007600572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,3072,0.01700799994998508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,3584,0.04370222157902188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,2560,0.014819555812411837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,3072,0.041138665543662176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,3072,0.022107554806603327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,2048,0.013020444247457715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,2560,0.03846399982770284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,2560,0.020917332834667627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,1536,0.010580444501505958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,2048,0.018927110566033255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,2048,0.03565066721704271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,1024,0.00832177781396442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,1536,0.0174666659699546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,1536,0.03286133209864298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,1024,0.015538665983412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,768,0.007524444825119442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,1024,0.030117332935333252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,512,0.005660444498062134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,768,0.014491554763582019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,768,0.02796799937884013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,256,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,512,0.013797333670987023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,512,0.025794666674402025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,128,0.003954666770166821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,256,0.024051555328898962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,256,0.012853333519564735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,128,0.012843555874294706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,64,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,12288,32,0.004218666503826777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,12288,128,0.021998221675554912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,64,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,12288,32,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,65536,0.2412924501630995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,65536,0.25271378623114693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,51200,0.19008444415198433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,51200,0.20366488562689888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,65536,0.367880900700887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,16384,0.06801511181725396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,16384,0.0731511116027832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,51200,0.29175644450717503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,12288,0.050445334778891675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,16384,0.10954222414228652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,12288,0.05779822005165947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,10240,0.04334577918052673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,12288,0.0880977776315477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,8192,0.03649333450529311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,10240,0.05024977856212192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,10240,0.07807555463578966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,7168,0.031231999397277832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,8192,0.06818399826685588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,7168,0.036730666955312095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,6144,0.02780533168050978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,7168,0.06150844362046984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,6144,0.03302399979697333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,6144,0.05675111214319865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,5120,0.02420711186197069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,5120,0.028208888239330713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,5120,0.05168799890412224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,4096,0.019872888922691345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,8192,0.042796442906061806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,3584,0.01757777730623881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,4096,0.04544088906712002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,4096,0.025352888637118872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,3072,0.015536889433860779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,3584,0.04248799880345663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,3072,0.040045334233178034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,3072,0.022058667408095464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,2560,0.01370133293999566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,2560,0.03726844324005975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,2560,0.020289777053727042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,2048,0.01163111130396525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,2048,0.03504800134234958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,2048,0.018731556004948087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,1536,0.00980444418059455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,1536,0.016911110944218107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,1024,0.008282666405042013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,3584,0.02310133311483595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,1024,0.029462221595976088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,1024,0.014987554815080432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,768,0.006647999915811751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,768,0.02777333392037286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,768,0.014183110660976835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,512,0.005413333161009683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,512,0.02540622154871623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,256,0.004266666869322459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,512,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,1536,0.03295200069745382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,256,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,256,0.023795555035273235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,128,0.004384888956944148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,64,0.0038648889296584656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,128,0.012505778008037143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,10240,128,0.021996445126003687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,10240,32,0.0038782221575578055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,64,0.011786667009194693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,65536,0.18044800228542754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,10240,32,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,65536,0.21325955126020643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,51200,0.14085955089992946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,51200,0.17045956187778047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,16384,0.05143644412358602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,16384,0.062485330634646945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,65536,0.24239555994669595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,12288,0.03836444351408217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,51200,0.19537599881490073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,16384,0.08052177561653985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,10240,0.03313244382540385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,12288,0.04974311259057787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,12288,0.06555289030075073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,10240,0.043176889419555664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,10240,0.05831111139721341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,8192,0.03565866748491923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,8192,0.05119644602139791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,7168,0.02404888967672984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,7168,0.04690133200751411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,7168,0.031297776434156634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,6144,0.02176800039079454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,6144,0.02783999840418498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,6144,0.04367555512322319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,5120,0.018583999739752877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,5120,0.024937777055634394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,5120,0.0396088891559177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,4096,0.0155102229780621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,8192,0.02791644467247857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,3584,0.013765333427323235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,4096,0.022456889351209004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,4096,0.03663377629386054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,3072,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,3584,0.021317332983016968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,3584,0.034790221187803484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,3072,0.019880000087949965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,2560,0.011031111081441244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,3072,0.03380799955791897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,2560,0.018418666389253404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,2560,0.032497776879204646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,2048,0.009763555394278632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,2048,0.017278222574128043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,2048,0.03087555699878269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,1536,0.00832444429397583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,1536,0.015545777148670621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,1536,0.027460444304678176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,1024,0.006299555715587404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,1024,0.014105778601434497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,768,0.005613333235184352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,768,0.013791110780504016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,768,0.025436444414986506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,512,0.004297777596447202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,512,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,512,0.023678221636348303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,256,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,1024,0.0248026665714052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,256,0.0123848890264829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,256,0.021976888179779053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,128,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,128,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,8192,128,0.02089244458410475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,64,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,8192,32,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,64,0.01240355521440506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,8192,32,0.012910222013791403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,65536,0.16850577460394967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,65536,0.2095573345820109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,51200,0.13139732678731283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,51200,0.16762222184075248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,65536,0.2381502257453071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,16384,0.04822577701674568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,16384,0.061477336618635386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,51200,0.19121689266628691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,12288,0.03566844595803155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,16384,0.07884177896711561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,12288,0.04884711239073011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,12288,0.06429155667622884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,10240,0.03113244308365716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,10240,0.042371554507149585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,10240,0.05754488706588745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,8192,0.026183111800087824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,8192,0.0506257779068417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,7168,0.022475555539131165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,7168,0.030574222405751545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,7168,0.046037332879172436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,6144,0.020030222005314298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,6144,0.02717777755525377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,6144,0.04299022091759575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,8192,0.03446844551298354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,5120,0.017264889346228707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,5120,0.02444711161984338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,5120,0.03933866818745931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,4096,0.014364444547229342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,4096,0.02201333310869005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,4096,0.03607733382119073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,3584,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,3584,0.020786666207843356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,3584,0.03432888785998026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,3072,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,3072,0.01982933282852173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,3072,0.03356444504525926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,2560,0.010272888673676385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,2560,0.018434666925006442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,2560,0.03188177943229675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,2048,0.009381333159075843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,2048,0.016556445095274184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,1536,0.008033778104517195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,1536,0.015311999453438653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,1536,0.026739555928442214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,1024,0.005710222240951326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,1024,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,1024,0.024824889169798955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,768,0.005904888941182031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,2048,0.03022222386466132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,768,0.013418667018413544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,512,0.0048942222363419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,768,0.024844444460339014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,512,0.013058666553762225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,512,0.022712000542216833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,256,0.004212444441186057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,256,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,256,0.02274399995803833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,128,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,128,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,7168,128,0.019336000084877014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,64,0.003289777785539627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,7168,32,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,64,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,65536,0.15753955311245388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,7168,32,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,65536,0.17378577921125624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,51200,0.12145155005984837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,51200,0.1397600041495429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,16384,0.0477804442246755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,65536,0.22514579031202528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,16384,0.05199466811286079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,51200,0.1810364458296034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,12288,0.03655111127429538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,16384,0.07417244381374784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,12288,0.04185511006249321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,12288,0.0610026650958591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,10240,0.031488888793521456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,10240,0.03536977701716953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,8192,0.02648711038960351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,10240,0.05515022079149882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,8192,0.029217776325013902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,7168,0.023079110516442194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,8192,0.049435555934906006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,7168,0.02569244470861223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,6144,0.02018222212791443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,6144,0.023777777949968975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,5120,0.01773955590195126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,6144,0.04189777705404493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,7168,0.04477689001295301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,4096,0.014202665951516895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,5120,0.021632888250880774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,5120,0.038463112380769514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,3584,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,4096,0.03547733359866672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,4096,0.019909333851602342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,3072,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,3584,0.03360355562633938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,3584,0.019111999207072787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,2560,0.010383110907342698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,3072,0.017924444542990792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,3072,0.032568001084857516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,2048,0.009104889300134447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,2560,0.01686311099264357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,2560,0.030791110462612573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,2048,0.015573332707087198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,1536,0.007658667034573025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,2048,0.02886577778392368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,1024,0.005659555395444234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,1536,0.014121777481502958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,1536,0.025846223036448162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,768,0.004949333353175057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,1024,0.024732444021436904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,1024,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,512,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,768,0.02442311081621382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,768,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,256,0.0036026665733920205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,512,0.012746666868527731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,512,0.022665777140193518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,128,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,256,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,256,0.022691556149058874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,128,0.011376889215575324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,64,0.003488000068399641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,6144,32,0.0032951111594835916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,6144,128,0.019113777412308585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,64,0.01145866678820716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,65536,0.14108267095353869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,6144,32,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,65536,0.1702959934870402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,51200,0.11117422580718994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,51200,0.13608177502950033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,16384,0.04467288984192742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,65536,0.2228648927476671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,51200,0.1787031094233195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,12288,0.03424177898301019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,16384,0.07341866360770331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,12288,0.038887110021379255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,10240,0.029564443561765883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,12288,0.060179558065202504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,10240,0.033497777250077986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,8192,0.024400000770886738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,10240,0.05443377627266777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,16384,0.05010133319430881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,8192,0.028712888558705647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,7168,0.021734222769737244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,8192,0.04828088813357883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,6144,0.01887022289964888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,7168,0.04405422343148125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,7168,0.025669333007600572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,5120,0.015950222810109455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,6144,0.023539554741647508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,6144,0.04158844550450643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,4096,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,5120,0.02182666626241472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,5120,0.03744533326890733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,3584,0.012251555919647217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,4096,0.019907555646366544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,4096,0.035038222869237266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,3072,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,3584,0.03327911098798116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,3584,0.018623110320832994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,2560,0.010127999716334874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,3072,0.03247111042340597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,3072,0.01791466606987847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,2048,0.008632889224423302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,2560,0.030271109607484605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,2560,0.015847999188635085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,2048,0.014817777607176038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,1536,0.007260444263617198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,2048,0.027618666489919026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,1536,0.025771554973390367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,1536,0.013841778039932251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,1024,0.0052266667286554975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,1024,0.024105777343114216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,768,0.004980444494220945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,1024,0.012830221818553077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,768,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,512,0.004664888812435998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,768,0.023494222097926672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,512,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,256,0.003900444342030419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,512,0.023357333408461675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,256,0.01181688904762268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,128,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,256,0.02163999941613939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,128,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,5120,128,0.018941332896550495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,64,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,5120,32,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,32,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,5120,64,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,65536,0.1030435562133789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,65536,0.16306489043765598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,51200,0.07990488741132948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,65536,0.22144356038835314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,51200,0.13133866257137722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,16384,0.033816887272728816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,16384,0.04816533459557427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,51200,0.17764888869391549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,12288,0.027426666683620874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,16384,0.07314488622877333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,12288,0.03713422351413303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,10240,0.020199111766285367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,12288,0.05936444467968411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,10240,0.03205777870284186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,8192,0.016837333639462788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,10240,0.05408177773157755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,8192,0.027687999937269423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,7168,0.014759999182489185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,8192,0.04806844393412272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,7168,0.025137777129809063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,6144,0.013199999928474426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,7168,0.04380977816051907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,5120,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,6144,0.022943110929595098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,6144,0.04095022214783563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,4096,0.0096124443742964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,5120,0.037019554111692644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,5120,0.021408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,3584,0.009220444493823582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,4096,0.034862223598692156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,4096,0.019336000084877014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,3072,0.008395555946562026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,3584,0.03279733326699998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,3584,0.017638223038779367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,2560,0.007673777639865875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,3072,0.016640888320075143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,3072,0.03201777736345927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,2048,0.006616000086069107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,2560,0.029530667596393164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,2048,0.014313777287801107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,1536,0.0052693333062860705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,2048,0.027128888501061335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,1536,0.01348355578051673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,1024,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,1536,0.026424888107511733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,1024,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,1024,0.02478933334350586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,768,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,768,0.023984889189402264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,768,0.012840888566441007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,512,0.00378222225440873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,512,0.01239377756913503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,512,0.02238222294383579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,256,0.0032275555034478507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,2560,0.015202666322390238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,128,0.0033208889265855155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,256,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,256,0.022299556268586054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,64,0.0031644445326593188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,4096,128,0.018283555905024212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,4096,32,0.0029502221279674103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,64,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,65536,0.09317777554194133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,65536,0.162228438589308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,51200,0.07411111063427396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,65536,0.22032888730367026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,51200,0.13027111689249674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,16384,0.030510223574108545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,51200,0.17683111296759713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,4096,32,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,12288,0.0247715562582016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,16384,0.07258311245176527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,16384,0.04706755611631605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,12288,0.03628888726234436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,10240,0.023294222023752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,12288,0.059319999482896596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,10240,0.03195200032658047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,8192,0.015655999382336933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,10240,0.0534746679994795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,8192,0.027640889088312786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,8192,0.04782311121622721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,7168,0.024819556209776137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,7168,0.04342755675315857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,6144,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,6144,0.023111111587948267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,5120,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,7168,0.013811555173661975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,6144,0.04148711098564996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,4096,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,5120,0.037279999918407865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,4096,0.0191920002301534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,4096,0.03439555565516154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,3584,0.008592888712882996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,3584,0.016920000314712524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,3072,0.008232889076073965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,3584,0.031594667169782854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,3072,0.0158897770775689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,3072,0.0321999986966451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,2560,0.007393777370452881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,2560,0.027992000182469685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,5120,0.021399110555648804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,2048,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,2560,0.015507555670208402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,1536,0.005299555758635203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,2048,0.026750221848487854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,2048,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,1024,0.00425511101881663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,1536,0.025807110799683466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,1536,0.013831111292044321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,768,0.003947555604908201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,1024,0.02403733299838172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,1024,0.01293777757220798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,512,0.0036657779581016968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,768,0.023725334140989516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,768,0.012853333519564735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,512,0.022368000613318548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,512,0.012176888684431711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,256,0.01145155562294854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,256,0.021714667479197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,128,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3584,128,0.018381334013409086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,128,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,32,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,64,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,65536,0.07563110854890612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3584,32,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3584,256,0.0035822221802340615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,65536,0.12498044967651367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,51200,0.05992799997329712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,51200,0.10067822535832723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,65536,0.22016533215840658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,16384,0.02678488857216305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,51200,0.17745067013634575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,16384,0.07223822010887994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,12288,0.021108445194032457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,16384,0.0372613337304857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,10240,0.02150933278931512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,12288,0.029000888268152874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,12288,0.05940710835986667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,8192,0.01756088932355245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,10240,0.05365066727002462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,10240,0.026408000124825373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,7168,0.01623288955953386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,8192,0.023771555887328252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,8192,0.047298666503694325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,6144,0.014530667000346713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,7168,0.021010667085647583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,7168,0.04361600014898512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,6144,0.019367999500698514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,5120,0.012607110871209038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,5120,0.017924444542990792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,5120,0.037548445992999606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,4096,0.016554666890038382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,4096,0.033687111404207014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,3584,0.00977600034740236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,3584,0.015818667080667283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,6144,0.0406951109568278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,4096,0.010612444745169746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,3072,0.00867555538813273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,3072,0.014175999495718213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,2560,0.007643555601437886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,3072,0.0309840010272132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,2560,0.013809777796268463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,2048,0.0069875551594628235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,2560,0.02885777751604716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,2048,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,1536,0.005624889085690181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,2048,0.026849778162108526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,1536,0.012823111481136747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,1024,0.004588444613748127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,3584,0.03126666612095303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,1024,0.012395555774370829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,768,0.0042462220622433555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,1024,0.024392889605628118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,768,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,768,0.023815999428431194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,512,0.003927999900446998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,512,0.012129777835475074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,1536,0.025215110845035974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,256,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,256,0.021723555194007024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,128,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,128,0.01105955574247572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,512,0.024078221784697637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,64,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,3072,128,0.019055111540688407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,3072,32,0.0029528888149393927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,64,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,65536,0.06806488831837972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,3072,32,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,65536,0.12228533956739639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,51200,0.054670221275753446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,65536,0.2200817796919081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,16384,0.024140444066789415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,51200,0.09859822193781535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,16384,0.035383111900753446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,51200,0.17656355433993867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,16384,0.07258133093516032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,12288,0.01919822229279412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,10240,0.017532444662517972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,12288,0.059188445409139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,12288,0.02884888980123732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,8192,0.014944000376595391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,10240,0.026111111044883728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,10240,0.05375466744105021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,7168,0.013660444153679742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,8192,0.047712001535627574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,8192,0.023175999522209167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,6144,0.01221244451072481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,7168,0.020820443828900654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,7168,0.04310399956173367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,5120,0.010802666346232096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,6144,0.019077334139082167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,6144,0.04083199964629279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,5120,0.01793155570824941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,4096,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,5120,0.03703377644220988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,4096,0.01588622232278188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,3584,0.009711999860074785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,3584,0.014846222268210517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,3072,0.008103110724025303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,3072,0.014180445008807711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,3072,0.029882666137483384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,2560,0.006779555645253923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,2560,0.013487999637921652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,4096,0.03497688968976339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,2048,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,2560,0.028562668297025893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,3584,0.030926220946841772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,1536,0.005715555614895291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,2048,0.01349688900841607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,2048,0.026844445202085707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,1024,0.004947555561860402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,1536,0.02565777798493703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,1024,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,768,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,768,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,1024,0.024327110913064744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,512,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,768,0.02409511142306858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,512,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,256,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,512,0.023039110832744177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,256,0.022376888328128394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,256,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,128,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,1536,0.012468444804350534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2560,128,0.01899111105336083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,128,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,64,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2560,32,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,65536,0.056587556997934975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,32,0.011335111326641507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,65536,0.11842044194539388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,51200,0.04470400015513102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,65536,0.2195440000957913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,51200,0.09581955273946126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,16384,0.019111111760139465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2560,64,0.011033777561452655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,16384,0.035084442959891424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,12288,0.016037333342764113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,12288,0.02849510974354214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,12288,0.059233778052859835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,10240,0.015078221758206686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,51200,0.17645511362287733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,10240,0.02585244509908888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,8192,0.01349066694577535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,16384,0.07159555620617337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,10240,0.05309155583381653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,8192,0.02312177750799391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,7168,0.012857777376969656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,8192,0.04753511150677999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,6144,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,7168,0.04313244422276815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,7168,0.020388444264729817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,6144,0.01891999940077464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,5120,0.009910222556855943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,6144,0.04074666566318936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,5120,0.016844444804721408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,4096,0.008445333275530074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,5120,0.03732000125779046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,4096,0.015542222393883599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,3584,0.0076248885856734375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,4096,0.03367111086845398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,3584,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,3584,0.03204088740878635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,3072,0.006892444358931647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,3072,0.029845333761639063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,2560,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,3072,0.014668444792429606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,2560,0.013423999978436364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,2560,0.02847111225128174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,2048,0.013165333204799227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,1536,0.004966222163703707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,2048,0.027074666486846075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,1536,0.012445333103338877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,1024,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,1024,0.012056888805495368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,2048,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,768,0.0038480001191298165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,1024,0.02372177773051792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,768,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,768,0.023397333092159692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,512,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,512,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,512,0.02275111112329695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,256,0.0032026666320032547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,256,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,256,0.020996444755130347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,128,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,1536,0.025765332910749648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,2048,128,0.018975999620225694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,128,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,2048,32,0.0028168888141711554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,65536,0.0437102218468984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,64,0.011517333487669626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,2048,32,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,51200,0.034632888105180525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,65536,0.11578222115834554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,51200,0.09133599864112006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,16384,0.016175111134847004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,65536,0.22031022442711723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,51200,0.17632710933685303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,12288,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,16384,0.03398044572936164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,12288,0.058598220348358154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,10240,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,16384,0.07218399975034925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,10240,0.025369778275489807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,10240,0.05336088935534159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,8192,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,8192,0.04712355468008253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,8192,0.022304889228608873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,7168,0.01187999960449007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,7168,0.019894222418467205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,7168,0.042542222473356456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,6144,0.011000889043013254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,6144,0.01793066660563151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,5120,0.009375111096435124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,6144,0.03874222106403775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,12288,0.027842667367723253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,4096,0.00795111142926746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,5120,0.016315556234783597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,5120,0.03532888823085361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,3584,0.007343110938866933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,4096,0.015457777513398064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,4096,0.031919111808141075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,3072,0.006551110910044775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,3584,0.014833778142929077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,3584,0.03020177616013421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,2560,0.005602666487296422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,3072,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,3072,0.029543999168607924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,2048,0.005264889034960005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,2560,0.013419555293189155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,2560,0.027492443720499676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,1536,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,2048,0.02678577767478095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,2048,0.01312888910373052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,1024,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,1536,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,1536,0.02625244524743822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,768,0.0038542221817705366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,1024,0.012472000386979846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,1024,0.024039111203617517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,512,0.003527111063400904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,768,0.012038222617573209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,768,0.023355555203225877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,256,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,512,0.011700444751315646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,512,0.022686221533351477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,128,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,256,0.02107200026512146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,256,0.011693333586057028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,128,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,64,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1536,128,0.0203813330994712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,64,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1536,32,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,65536,0.030277334981494482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1536,32,0.011360000405046674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,65536,0.10933778021070693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,51200,0.02573866645495097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,65536,0.22007555431789824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,51200,0.0880933337741428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,16384,0.016937777400016785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,51200,0.1766151057349311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,16384,0.07156622409820557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,16384,0.033942222595214844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,12288,0.013834666874673633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,12288,0.027481777800454035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,12288,0.05807199743058947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,10240,0.012859555582205454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,10240,0.024498666326204937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,8192,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,10240,0.053136888477537364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,8192,0.021555556191338435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,8192,0.04719644453790453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,7168,0.010774222513039907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,7168,0.041590223709742226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,6144,0.009356444080670675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,6144,0.01751377847459581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,6144,0.03840533230039809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,5120,0.008646222452322641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,5120,0.016355555918481614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,5120,0.03534133235613505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,7168,0.019253333409627277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,4096,0.015483554866578845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,4096,0.03275200062327915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,3584,0.010509333676762052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,3584,0.014633778068754407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,4096,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,3072,0.009408888717492422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,3584,0.031062222189373438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,3072,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,2560,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,3072,0.029507554239696924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,2560,0.013447999954223633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,2048,0.007368889119890001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,2560,0.02844711144765218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,2048,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,2048,0.027107555005285475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,1536,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,1536,0.026127111580636766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,1536,0.012851555314328937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,1024,0.004459555364317364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,768,0.003972444269392225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,1024,0.024225778049892847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,768,0.01183377785815133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,768,0.023443554838498432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,512,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,512,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,512,0.022361778550677832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,256,0.011552888486120435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,256,0.021436444587177698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,128,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,1024,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,128,0.011231111155615913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,64,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,1024,128,0.01995377739270528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,1024,32,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,65536,0.02703999976317088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,64,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,1024,32,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,65536,0.11167555385165745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,51200,0.021091555555661518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,51200,0.08925777673721313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,65536,0.2204115523232354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,16384,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,51200,0.1773137781355116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,12288,0.01332711097266939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,16384,0.07116444243325128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,16384,0.03363110952907138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,10240,0.012169777519173093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,12288,0.026741334133678015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,12288,0.05847733550601535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,8192,0.010374222364690568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,10240,0.02324177821477254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,10240,0.05279644330342611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,8192,0.020938666330443487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,7168,0.009776888622177972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,7168,0.018908444378111098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,8192,0.044588443305757314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,6144,0.009343999955389235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,7168,0.04272266560130649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,5120,0.008319999608728621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,6144,0.03782400157716539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,6144,0.017687110437287223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,4096,0.008101333346631791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,5120,0.016855110724767048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,5120,0.03552711009979248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,3584,0.006931555353932911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,4096,0.01532799998919169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,4096,0.03298133280542161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,3584,0.030928889910380047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,3584,0.014527999692493014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,3072,0.006594666590293248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,3072,0.014318222800890604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,3072,0.03005511230892605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,2560,0.0057297779454125305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,2560,0.013567110730542077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,2048,0.005168000029193031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,2560,0.027691556347741023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,2048,0.013129777378506131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,1536,0.004618666652176115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,2048,0.026514665948020086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,1536,0.024872889121373493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,1024,0.004023999803596073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,1024,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,1024,0.02344977855682373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,768,0.003934222377008862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,1536,0.012764444781674279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,768,0.01184800018866857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,512,0.003640888879696528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,768,0.023784889115227595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,256,0.0031724443866146933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,512,0.022446221775478784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,512,0.011608888705571493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,128,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,256,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,256,0.021091555555661518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,128,0.011233777635627322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,768,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,768,128,0.01997511088848114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,65536,0.020232000284724765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,64,0.010837333069907295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,768,32,0.010886222124099731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,51200,0.022379555635982092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,65536,0.11077155669530232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,65536,0.22019644578297934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,51200,0.08810666534635757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,16384,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,51200,0.17711199654473198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,16384,0.032469332218170166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,12288,0.010976889067225985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,16384,0.0707884430885315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,10240,0.009927111367384592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,12288,0.025891555680169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,12288,0.05728089147143894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,8192,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,10240,0.050916443268458046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,10240,0.023128888673252527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,7168,0.008072000410821702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,8192,0.04462488823466831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,8192,0.02104533380932278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,6144,0.007701333198282454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,7168,0.040821333726247154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,7168,0.018739556272824604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,5120,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,6144,0.01758311192194621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,6144,0.037871999873055354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,4096,0.007666666474607255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,5120,0.03496444556448195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,5120,0.01626844373014238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,4096,0.015132443772421943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,3584,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,4096,0.03257511059443156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,3584,0.014125333891974555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,3072,0.006276444428496891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,3584,0.02991822361946106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,2560,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,3072,0.013853333062595792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,3072,0.02948533495267232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,2048,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,2560,0.013174222575293647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,2048,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,1536,0.004589333302444882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,2048,0.026514665948020086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,1536,0.012757333616415659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,1536,0.024895111719767254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,1024,0.003923555629120933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,1024,0.024373332659403484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,1024,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,768,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,2560,0.02715377840730879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,768,0.023456000619464453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,768,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,512,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,512,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,512,0.022575111852751836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,256,0.011148444480366178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,128,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,256,0.02202755543920729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,128,0.010666666759385003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,64,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,512,32,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,512,128,0.01871200071440803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,64,0.010672888822025724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,512,32,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,65536,0.10638133022520278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,65536,0.22071644994947645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,51200,0.015249777171346875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,51200,0.08672889073689778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,65536,0.017640888690948486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,16384,0.00979555563794242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,51200,0.17699733045366076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,16384,0.03144177794456482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,16384,0.0711866683430142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,12288,0.008310221963458592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,12288,0.05555999941296048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,12288,0.026127111580636766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,10240,0.013085333009560903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,10240,0.023429332507981196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,10240,0.05009955498907301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,8192,0.010795555180973478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,8192,0.04491466614935133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,8192,0.020764445265134174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,7168,0.010072000324726105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,7168,0.04082666503058539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,7168,0.018632888793945312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,6144,0.007668444679843054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,6144,0.038541333542929754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,6144,0.017422222428851657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,5120,0.00795911086930169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,5120,0.035058667262395225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,5120,0.01625066664483812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,4096,0.007616000043021307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,4096,0.014886221951908536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,4096,0.032285332679748535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,3584,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,3584,0.014554666148291694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,3584,0.030151999659008447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,3072,0.0069360000391801195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,2560,0.005675555517276128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,3072,0.029525332980685767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,2560,0.013591999808947245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,2560,0.02849600050184462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,2048,0.005010666532648934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,2048,0.013095999757448832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,2048,0.026553778184784785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,1536,0.004573333180612988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,1536,0.025384889708624944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,1024,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,3072,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,1024,0.012295111185974546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,768,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,1024,0.023443554838498432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,768,0.011968889170222811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,512,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,768,0.022774222824308608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,512,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,256,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,512,0.02241777711444431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,1536,0.012533333566453723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,128,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,256,0.011500444677140979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,256,0.02122666603989071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,128,0.011296889020336999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,256,128,0.01994044416480594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,32,0.002699555622206794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,64,0.010771555205186209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,65536,0.014429334137174817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,256,32,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,65536,0.10760177506340875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,51200,0.012990222209029727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,65536,0.2206453349855211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,51200,0.08493866523106892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,16384,0.009342222577995723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,256,64,0.00254133323000537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,51200,0.17706933286454943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,12288,0.008281778130266402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,16384,0.030972444348865088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,16384,0.06880088647206624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,10240,0.008736000292830998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,12288,0.05617244376076592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,12288,0.025669333007600572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,8192,0.007986666427718269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,10240,0.022822222775883142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,10240,0.05005955696105957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,7168,0.007654222349325816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,8192,0.020662221643659804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,8192,0.044518222411473594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,6144,0.007282666862010956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,7168,0.04099644554985894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,6144,0.01718311177359687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,6144,0.038463112380769514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,5120,0.006648889018429651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,5120,0.01649777756796943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,5120,0.034421334664026894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,4096,0.006587555425034628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,4096,0.032567110326555036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,4096,0.014858666393491956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,3584,0.006251555350091722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,7168,0.01864622202184465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,3072,0.006017777654859755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,3584,0.030941334035661485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,3584,0.01421511173248291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,2560,0.006247111078765657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,3072,0.013776000175211163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,3072,0.029534220695495605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,2560,0.027854220734702215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,2560,0.013169777890046438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,2048,0.006226666685607698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,2048,0.02641155487961239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,1536,0.0045422220395671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,2048,0.012704888979593912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,1536,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,1024,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,1536,0.02481599979930454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,1024,0.011843555503421359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,768,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,1024,0.02346399923165639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,768,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,768,0.02292088833120134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,512,0.0031822222388452957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,512,0.022703111171722412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,512,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,128,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,256,0.021000888612535264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,128,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,64,0.0025475554996066624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,160,128,128,0.018571555614471436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,128,32,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,64,0.010777778095669217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,65536,0.014485332700941296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,128,32,0.010830221904648675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,51200,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,65536,0.10600622495015462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,16384,0.008990222381220924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,51200,0.08428621954388088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,16384,0.03122133347723219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,12288,0.007714666426181793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,10240,0.007429333196745978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,8192,0.0069511110583941145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,12288,0.025519111090236243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,7168,0.006509333435032103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,10240,0.022857778602176245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,6144,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,8192,0.020478222105238173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,5120,0.0059075554211934405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,6144,0.01718933383623759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,4096,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,5120,0.0162435554795795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,4096,0.014885332849290637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,3584,0.006317333214812809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,3072,0.005941333456171884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,3584,0.014188443620999655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,2560,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,3072,0.013816888961527081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,2048,0.005954666684071223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,2560,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,1536,0.004547555413511065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,7168,0.018569777409235638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,2048,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,1536,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,1024,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,768,0.003489777859714296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,512,0.003269333392381668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,768,0.011786667009194693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,1024,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,256,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,512,0.011346666349305047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,128,0.0025155554629034465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,64,0.002546666603949335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,128,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,256,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,64,32,0.002519999941190084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,64,0.010696888797812991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,65536,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,51200,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,64,32,0.010753778119881948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,16384,0.008019555773999956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,65536,0.10516533586714004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,12288,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,51200,0.08331911431418525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,10240,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,16384,0.031123555368847314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,12288,0.02539377742343479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,8192,0.006605333338181178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,10240,0.022659555077552795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,7168,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,6144,0.005982222242487802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,8192,0.020426667398876615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,5120,0.006220444622966979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,6144,0.01734133395883772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,5120,0.016182222300105624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,4096,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,3584,0.006272000157170826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,4096,0.015140444040298462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,3072,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,3584,0.014150222142537435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,2560,0.005912888795137405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,3072,0.013743110828929476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,2048,0.00590844452381134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,2048,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,1536,0.004554666578769684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,2560,0.013185777597957186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,1024,0.003843555433882607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,7168,0.018233777748213876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,1536,0.012383999923865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,768,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,1024,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,512,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,768,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,256,0.0027955555253558685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,512,0.011355555719799466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,128,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,256,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,64,0.0025155554629034465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,128,0.010425777898894416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,160,32,32,0.0026346666531430352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,64,0.010684444672531553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,160,32,32,0.010386666489972008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,51200,0.8316168785095215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,16384,0.35567111439175075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,16384,0.26561421818203396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,12288,0.25676356421576607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,51200,1.022855970594618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,10240,0.21593156125810412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,12288,0.2029555506176419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,8192,0.17798844973246256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,10240,0.1719893349541558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,8192,0.13672178321414524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,7168,0.15128533045450845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,6144,0.13411466280619302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,7168,0.12218666076660156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,5120,0.11187822288937038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,6144,0.10711910989549424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,4096,0.09203555848863389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,5120,0.09143466419643825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,3584,0.08094133271111383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,4096,0.07601066430409749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,3072,0.07062488794326782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,3584,0.067958222495185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,2560,0.060158219602372914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,3072,0.06070488691329956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,2048,0.050712890095180936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,2560,0.05297955539491442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,1536,0.03914133376545376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,2048,0.045744889312320285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,1024,0.027980443504121568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,1536,0.03843644592497084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,768,0.022795556320084467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,1024,0.03118844495879279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,512,0.0159262220064799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,768,0.037420445018344455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,256,0.01037599974208408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,512,0.0288426677385966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,128,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,256,0.021376889612939622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,64,0.007824888659848107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,64,0.019609777463806998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,65536,32,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,32,0.027098667290475633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,65536,1.074737760755751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,65536,0.842264864179823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,65536,128,0.020089776979552377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,16384,0.2872524526384142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,16384,0.21239733695983887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,51200,0.6609084341261122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,51200,0.8472764227125379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,12288,0.21229510837131074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,10240,0.1967600054211087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,12288,0.16306132740444607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,8192,0.1629519992404514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,10240,0.13757333490583631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,7168,0.09819822178946601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,8192,0.11024444633060032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,7168,0.13890222708384195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,6144,0.1220853328704834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,6144,0.0865395532713996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,4096,0.08229511313968234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,5120,0.0743617746565077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,3584,0.07264088922076754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,4096,0.0622239973809984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,3072,0.06312800116009183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,3584,0.056206219726138644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,2560,0.05403911074002584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,3072,0.05044710967275831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,2048,0.04435555471314324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,2560,0.044603556394577026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,1536,0.034244444635179304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,2048,0.03898488812976413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,5120,0.10126133097542657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,1024,0.02478222217824724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,1536,0.033287998702791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,768,0.019498666127522785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,512,0.014486221803559197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,1024,0.027051554785834417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,768,0.0319164428446028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,512,0.02440533373090956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,128,0.00665155549844106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,256,0.018990221950742934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,64,0.005254222287072076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,128,0.017909333109855652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,32,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,64,0.017864889568752713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,51200,32,0.023404444257418316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,65536,0.33970488442315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,51200,256,0.009927111367384592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,65536,0.3346586757236057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,51200,0.2663244406382243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,16384,0.09522577789094712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,51200,0.26504890124003094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,12288,0.06889688968658447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,12288,0.07206222083833483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,16384,0.09090133508046468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,10240,0.058373332023620605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,8192,0.049729777706993945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,10240,0.06271466943952772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,7168,0.04212266537878248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,8192,0.04972355564435323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,6144,0.037504888243145414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,6144,0.04075377848413255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,5120,0.031643556223975286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,7168,0.045533332559797496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,4096,0.026358221968015034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,5120,0.03584266702334086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,3584,0.023362666368484497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,4096,0.03127022253142463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,3072,0.0203448881705602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,3584,0.027843554814656574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,2560,0.018138666947682697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,2048,0.015154666370815702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,2560,0.023381332556406658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,1536,0.011886222494973077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,1536,0.01940088967482249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,2048,0.02090399960676829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,1024,0.009381333159075843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,768,0.008393777741326226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,768,0.01793511046303643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,512,0.006421333385838403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,1024,0.016927111479971144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,256,0.004506666627195147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,3072,0.02611200014750163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,128,0.003912888881233004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,256,0.013551111022631327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,64,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,128,0.013463999662134381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,16384,32,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,64,0.012892444100644855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,32,0.01387466655837165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,16384,512,0.015972443752818637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,65536,0.2711724440256755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,51200,0.2113084395726522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,65536,0.25144799550374347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,16384,0.07644889089796278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,51200,0.20026666588253447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,12288,0.0554622213045756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,16384,0.07098311185836792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,12288,0.057075553470187716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,8192,0.03997600078582764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,10240,0.049843556351131864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,7168,0.034146666526794434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,8192,0.03988000088267856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,6144,0.030922667847739324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,7168,0.03657777773009406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,5120,0.026467555099063452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,6144,0.032814221249686346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,4096,0.022108443909221228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,10240,0.04744889007674324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,5120,0.02884977724817064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,3584,0.01919644408755832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,3072,0.016896888613700867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,4096,0.024938666158252295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,3584,0.02332533399264018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,3072,0.021643555826610986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,2048,0.012816000315878125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,2560,0.020635555187861126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,2048,0.01867644488811493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,1536,0.010613333019945355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,1024,0.008522666990756989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,1536,0.01753866672515869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,768,0.007635555333561367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,1024,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,512,0.0052577778697013855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,768,0.016872000363137987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,256,0.004592000196377437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,512,0.014918223023414612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,128,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,256,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,2560,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,64,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,12288,32,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,128,0.012969777815871768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,32,0.013182222015327878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,12288,64,0.012717333104875354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,65536,0.26064978705512154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,51200,0.2060533364613851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,65536,0.24311200777689615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,16384,0.07124088870154487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,12288,0.05380533470047844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,16384,0.06843199994828966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,10240,0.04571911030345493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,12288,0.0555573304494222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,8192,0.03860888878504435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,10240,0.048657778236601085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,7168,0.033294220765431724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,8192,0.038740442858801946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,7168,0.03542844454447428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,6144,0.02960444490114848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,51200,0.19286755720774332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,5120,0.025524444050259058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,6144,0.03184355629814996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,4096,0.02068711154990726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,5120,0.02889333168665568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,3584,0.01866755551762051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,4096,0.024254222710927326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,3072,0.01626844373014238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,3584,0.023144889209005568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,2560,0.014387556248241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,3072,0.021379555265108745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,2048,0.012512889173295764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,2560,0.02032533288002014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,1536,0.010416888528399997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,2048,0.01794933279355367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,1024,0.008238222036096785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,1536,0.016965332958433364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,768,0.006689777804745569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,1024,0.014292443792025248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,768,0.016211556063758004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,512,0.0053084443012873335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,256,0.004615111069546806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,512,0.014515555567211576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,128,0.00397155558069547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,256,0.012865777644846173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,64,0.003670222229427762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,10240,32,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,128,0.012484444512261285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,64,0.012168889244397482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,10240,32,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,65536,0.19158132870992026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,51200,0.15075555112626818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,16384,0.05394666724734836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,65536,0.20142043961418998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,12288,0.040186666780047946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,51200,0.1610942284266154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,16384,0.05792178048027886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,10240,0.034824000464545354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,8192,0.02920088834232754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,10240,0.041482667128245033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,7168,0.02533422244919671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,12288,0.047319110896852284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,6144,0.0226248883538776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,8192,0.03312622176276313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,5120,0.019299555155966017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,7168,0.030054221550623577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,4096,0.015803555647532146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,6144,0.02742133372359806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,3584,0.014167999227841696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,5120,0.024504888388845656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,4096,0.02165599995189243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,3072,0.01276266657643848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,2560,0.011534222298198275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,3072,0.01928622192806668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,3584,0.020656888683636982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,2048,0.010089777410030365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,1536,0.008587555752860175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,2560,0.018592000007629395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,1024,0.0063680000603199005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,1536,0.015252444479200574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,2048,0.0170133329100079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,768,0.00509333320789867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,512,0.004621333132187526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,768,0.015239111251301236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,512,0.01349244432316886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,1024,0.013982221484184265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,256,0.003932444585694207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,128,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,64,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,256,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,8192,32,0.0035662220584021676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,128,0.011865778101815118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,64,0.012335110869672565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,8192,32,0.013119999733236102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,65536,0.16573244995541042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,51200,0.1299253304799398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,65536,0.1978515519036187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,16384,0.05179466803868612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,12288,0.03894933395915561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,51200,0.15823022524515787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,16384,0.058127111858791776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,10240,0.03366577625274658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,12288,0.04679200053215027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,8192,0.028199109766218398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,7168,0.024527110987239416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,10240,0.040656887822681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,8192,0.03235466612709893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,6144,0.02164622147878011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,5120,0.018595554762416415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,7168,0.02898400028546651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,6144,0.026312000221676294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,4096,0.015341333217091031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,3584,0.013805333111021252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,5120,0.024062222904629175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,4096,0.02125155594613817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,3072,0.012494222157531314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,2560,0.011173333558771344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,3072,0.019291554888089497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,3584,0.02069155540731218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,2048,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,2560,0.01809599995613098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,1536,0.008314666648705801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,2048,0.016354666815863717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,1024,0.005942222144868638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,1536,0.015088000231319003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,768,0.004914666629499859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,1024,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,768,0.014217777384652032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,512,0.00433955548538102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,256,0.003998222036494149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,512,0.013346666263209449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,128,0.0035004446076022256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,256,0.012156444291273752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,64,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,64,0.011805333197116852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,7168,32,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,32,0.011503111157152386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,65536,0.15334578355153403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,7168,128,0.011728888584507836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,65536,0.16096799903445774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,51200,0.11928178204430474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,16384,0.04536088969972399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,51200,0.1285946634080675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,12288,0.03429955575201247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,16384,0.048476444350348584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,10240,0.029606223106384277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,10240,0.03308355477121141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,12288,0.03897510965665182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,8192,0.02683288852373759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,7168,0.021386666430367365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,7168,0.025179555018742878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,6144,0.01885688967174954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,6144,0.02295733326011234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,8192,0.02476355599032508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,5120,0.01623644431432088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,4096,0.013345777988433838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,5120,0.0209840006298489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,3584,0.012166221936543783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,4096,0.01928711103068458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,3072,0.011031111081441244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,3584,0.01828444500764211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,2560,0.009860444400045607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,3072,0.017645332548353408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,2560,0.0162773331006368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,2048,0.008641777767075432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,1536,0.007299555672539606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,2048,0.014827556080288358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,1024,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,768,0.00488711107108328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,1536,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,512,0.00425511101881663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,1024,0.013198222551080914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,768,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,256,0.003682666768630346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,128,0.0034577778230110803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,256,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,64,0.003104888937539525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,512,0.012868444124857584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,6144,32,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,128,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,64,0.011703111231327057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,65536,0.119687106874254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,6144,32,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,51200,0.09534222549862331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,16384,0.039176887936062284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,65536,0.1575066645940145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,12288,0.03232266505559286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,51200,0.12564533286624485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,10240,0.027442667219373915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,16384,0.0469760000705719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,12288,0.03653511073854234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,8192,0.02271555529700385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,7168,0.02015822297996945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,8192,0.026216889421145122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,6144,0.017542221479945712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,10240,0.031509333186679415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,7168,0.024858666790856257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,5120,0.014910222755538093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,4096,0.012634666429625617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,6144,0.022337777747048274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,3584,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,5120,0.020940444535679285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,3072,0.010425777898894416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,4096,0.01909955508179135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,3584,0.018114666144053142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,2560,0.009343999955389235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,3072,0.01684711045689053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,2048,0.008387555678685507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,2560,0.015751111838552687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,1536,0.0069520001610120135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,1024,0.0052764444715446895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,2048,0.014456000592973499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,768,0.004555555681387584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,1536,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,512,0.004192000048028098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,1024,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,768,0.013890667094124688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,256,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,512,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,128,0.003485333174467087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,256,0.011832888755533429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,64,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,128,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,5120,32,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,64,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,5120,32,0.011901333100265928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,65536,0.10712355375289917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,51200,0.08341066704856025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,65536,0.1517875591913859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,16384,0.03405155407057868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,16384,0.04468355576197306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,51200,0.12206044461992051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,12288,0.027526222997241553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,10240,0.02202577723397149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,12288,0.03532622257868449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,10240,0.030861334668265447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,8192,0.017820444371965196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,7168,0.015692444311247934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,8192,0.02566844390498267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,6144,0.014146667387750415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,7168,0.024130667249361675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,5120,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,6144,0.022433777650197346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,4096,0.010464888479974536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,5120,0.020568889048364427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,3584,0.009535999761687385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,4096,0.018375110295083787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,3072,0.0085235552655326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,2560,0.007920000288221572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,3584,0.017350221673647564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,3072,0.01624000072479248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,2048,0.006766222417354584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,1536,0.0052879999081293745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,2048,0.014183110660976835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,1024,0.004600000050332812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,2560,0.01551644504070282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,768,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,1536,0.013552889227867126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,1024,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,512,0.003626666549179289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,256,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,512,0.012834666503800286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,768,0.013602666556835175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,128,0.0032337777730491427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,64,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,256,0.01183555523554484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,4096,32,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,64,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,128,0.011696889168686338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,4096,32,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,65536,0.09170489178763495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,51200,0.07027466429604425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,16384,0.029089778661727905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,51200,0.12089778317345513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,65536,0.1506755616929796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,12288,0.024007999234729346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,10240,0.023558222585254248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,16384,0.043050666650136314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,8192,0.016883555385801528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,10240,0.030795554320017498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,7168,0.014936000108718872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,8192,0.02574133376280467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,6144,0.013169777890046438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,7168,0.023870221442646448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,5120,0.011976000335481433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,6144,0.022240888741281297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,5120,0.02032711108525594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,4096,0.010069333016872406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,3584,0.009363555245929295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,4096,0.017921777235137094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,3072,0.00870311094654931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,3584,0.016978666186332703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,2560,0.006990222467316522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,3072,0.015692444311247934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,12288,0.034713778230879046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,2048,0.006320888797442119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,1536,0.005252444495757421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,2560,0.014934221903483072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,1024,0.004233777936961916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,1536,0.013805333111021252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,2048,0.014504889647165934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,768,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,512,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,1024,0.012857777376969656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,512,0.012511110968059964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,256,0.003504000190231535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,768,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,128,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,64,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,256,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3584,32,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,128,0.011415111521879831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,64,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,65536,0.07869244284100003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3584,32,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,51200,0.0628382232454088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,16384,0.02700888944996728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,51200,0.09157244364420573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,65536,0.1131697760687934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,12288,0.021160889003011916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,16384,0.03350311186578538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,10240,0.019446222318543326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,12288,0.027560888065232172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,8192,0.014217777384652032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,10240,0.025026667449209426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,7168,0.01296533313062456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,8192,0.021376889612939622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,6144,0.011586666935020022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,7168,0.019945777124828763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,5120,0.010440000229411654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,6144,0.0188391109307607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,5120,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,4096,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,3584,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,4096,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,3072,0.007962666451931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,3584,0.014575110541449653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,2560,0.007078222102589077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,3072,0.014115555418862237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,2048,0.005994666781690385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,2560,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,2048,0.013059555656380124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,1536,0.013058666553762225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,1024,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,1024,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,768,0.004257777912749184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,768,0.012551999754375882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,512,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,512,0.012409778104888068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,256,0.003539555602603488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,1536,0.0052488889131281115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,256,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,64,0.003157333367400699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,128,0.01074577785200543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,3072,32,0.0029173334025674393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,64,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,65536,0.0674462252193027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,3072,32,0.011141333315107556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,51200,0.05378489030732048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,65536,0.10995466841591729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,16384,0.024567110670937434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,51200,0.08904621998469035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,12288,0.01959911154376136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,16384,0.03257777624660068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,10240,0.017873777283562552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,12288,0.02713599966631995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,10240,0.02459377712673611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,8192,0.0207022229830424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,7168,0.012247111234400006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,7168,0.019644444187482197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,6144,0.011052444577217102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,6144,0.017924444542990792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,5120,0.010073777702119615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,5120,0.016176000237464905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,4096,0.008761777646011775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,8192,0.013654222091039022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,3584,0.00831200016869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,4096,0.014807111687130399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,3072,0.007465778125656976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,2560,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,3584,0.01457244488928053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,2048,0.005658666706747479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,3072,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,1536,0.004884444591071871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,2048,0.013173333472675748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,1536,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,2560,0.013568888935777875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,1024,0.004275555411974589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,768,0.004100444416205089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,1024,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,512,0.003664888855483797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,768,0.012585777375433179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,256,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,128,0.002985777747299936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,512,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,64,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,256,0.011425777441925473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2560,32,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,64,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,65536,0.059867554240756564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,128,0.010848888920413123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2560,32,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,51200,0.04625244604216682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,16384,0.018979556030697294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,65536,0.10703555742899577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,51200,0.08661066823535496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,12288,0.016813332835833233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,10240,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,16384,0.03166666626930237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,12288,0.02686311139000787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,8192,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,10240,0.02421155571937561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,7168,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,6144,0.01071733319097095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,8192,0.02067733307679494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,5120,0.009671111073758867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,7168,0.018789332773950364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,6144,0.017792888813548617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,4096,0.008405333591832055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,5120,0.01586844523747762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,3584,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,3584,0.014210666219393412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,3072,0.0069226668112807805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,3072,0.013742222554153867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,2560,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,2560,0.013535999589496188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,2048,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,2048,0.012844444149070315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,1536,0.005079999979999331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,1536,0.01276266657643848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,4096,0.014545778433481852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,1024,0.004257777912749184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,768,0.004150222159094281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,1024,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,768,0.012056888805495368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,512,0.0036159998012913596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,512,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,128,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,256,0.011253332926167382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,64,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,2048,32,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,128,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,65536,0.043284446001052856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,64,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,2048,32,0.010779555473062726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,51200,0.0347031123108334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,65536,0.10516622331407334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,16384,0.01629333363638984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,12288,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,51200,0.08439555433061387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,16384,0.030953778160942927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,10240,0.012482666307025485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,8192,0.011326221956147088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,10240,0.02371999952528212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,7168,0.01092355532778634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,8192,0.019796444310082328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,6144,0.010335110955768162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,7168,0.01832444469134013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,5120,0.009665778113736046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,6144,0.016846223009957206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,5120,0.015875554747051664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,4096,0.008067555725574493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,3584,0.007351111206743453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,4096,0.014477334088749357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,3072,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,12288,0.026479111777411565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,3584,0.014194667339324951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,2560,0.005933333188295364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,2048,0.0052737775776121355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,2560,0.013760888742076026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,1536,0.004884444591071871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,2048,0.013146667016877068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,1024,0.004273777620659934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,1536,0.012521777715947894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,768,0.003956444561481476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,1024,0.012173333101802401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,512,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,768,0.012216889195972018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,3072,0.013610666824711693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,512,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,128,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,64,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,256,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1536,32,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,128,0.011352000137170156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,64,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,65536,0.03108711043993632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1536,32,0.011047999891969891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,51200,0.026073777013354834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,16384,0.01370133293999566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,65536,0.10153955883449978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,12288,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,16384,0.030994666947258845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,10240,0.01092088884777493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,12288,0.025783111651738484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,8192,0.010110222631030613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,10240,0.02308533257908291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,7168,0.009367999931176504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,8192,0.019546666079097323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,6144,0.010433778166770935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,7168,0.018259555101394653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,5120,0.008912000391218398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,51200,0.08146399921841092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,6144,0.0169804443915685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,4096,0.007628444168302748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,5120,0.015936889582210116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,3584,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,3072,0.006309333360857434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,4096,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,3584,0.014555555250909595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,2560,0.005674666828579373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,3072,0.013527111046844058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,2048,0.005636444522274866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,1536,0.00490399988161193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,2560,0.013850666582584381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,2048,0.0127697777416971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,1024,0.004010666575696734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,1536,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,1024,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,768,0.003863999827040566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,512,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,768,0.012162666353914471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,256,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,512,0.012068444656001197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,128,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,256,0.011376000112957425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,64,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,128,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,1024,32,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,32,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,65536,0.028367999527189467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,65536,0.10045066807005142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,51200,0.02345866627163357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,51200,0.08016000191370647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,1024,64,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,16384,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,12288,0.010071111222108206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,16384,0.030093332131703694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,10240,0.009525333013799455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,12288,0.025736000802781846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,8192,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,10240,0.021722666091389124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,7168,0.010448888772063784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,8192,0.019075555933846366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,6144,0.00906399968597624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,7168,0.017975999249352347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,5120,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,6144,0.01720888912677765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,5120,0.015653333730167814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,4096,0.007652444144090016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,3584,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,4096,0.014474666780895658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,3072,0.0063164445261160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,3584,0.014530667000346713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,2560,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,3072,0.013506666653686099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,2048,0.0052595556610160405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,2560,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,1536,0.00462666650613149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,2048,0.01275199982855055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,1024,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,1536,0.012829333543777466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,1024,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,768,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,512,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,768,0.012071111136012606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,256,0.003162666741344664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,512,0.01205599970287747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,128,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,256,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,64,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,128,0.011517333487669626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,768,32,0.0029884444342719186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,64,0.01108888867828581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,65536,0.022696000006463792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,51200,0.01827022267712487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,768,32,0.011495999991893768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,16384,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,65536,0.09886311160193549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,12288,0.010718222293588849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,51200,0.07872888776991102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,16384,0.029338667790095013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,10240,0.009742221898502773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,12288,0.02369955513212416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,8192,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,7168,0.008016000191370646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,10240,0.02198222279548645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,8192,0.018582221534517076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,6144,0.00758755538198683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,5120,0.00867911097076204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,7168,0.01793066660563151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,4096,0.007368889119890001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,6144,0.016724444097942777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,3584,0.0070053330726093715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,5120,0.015850666496488783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,3072,0.006266666783226862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,4096,0.014521777629852295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,2560,0.00590133335855272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,3584,0.013863999810483722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,2048,0.005197333378924264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,3072,0.013947554760509066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,1536,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,2560,0.013415111435784234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,2048,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,1024,0.003903111235962974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,768,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,1536,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,512,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,1024,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,256,0.003126222226354811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,768,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,128,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,512,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,64,0.0028231110837724474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,128,0.011200889117187925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,64,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,512,32,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,32,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,65536,0.018668444620238412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,65536,0.09656622012456258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,51200,0.015927111109097798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,512,256,0.011333333121405708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,16384,0.009933333430025313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,51200,0.0759235554271274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,12288,0.01168444421556261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,16384,0.0284586681260003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,10240,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,12288,0.023944000403086346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,8192,0.009677333136399588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,10240,0.021672889590263367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,7168,0.008982222113344405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,8192,0.01886311173439026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,6144,0.008381333616044786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,7168,0.017608000172509086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,5120,0.007999111380841997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,6144,0.016778666112158034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,4096,0.007256000406212277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,5120,0.015451555450757345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,3584,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,4096,0.014481777946154276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,3072,0.006239999913507038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,3584,0.013771555489963956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,2560,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,3072,0.013798221945762634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,2048,0.004905777672926585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,2560,0.012763555679056378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,1536,0.00452533322903845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,2048,0.012727110750145383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,1024,0.0038622220357259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,1536,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,768,0.003693333516518275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,1024,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,512,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,768,0.011734222372372946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,256,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,512,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,128,0.002817777709828483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,256,0.011015999648306104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,128,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,64,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,256,32,0.010614222122563256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,65536,0.014701333310869006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,65536,0.09508799844317967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,51200,0.014483556151390076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,51200,0.07384000221888225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,16384,0.009438222481144799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,16384,0.028061333629820082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,12288,0.009344889058007134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,12288,0.023535999986860488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,10240,0.00869333330127928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,10240,0.021292444732454088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,8192,0.007983999947706858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,8192,0.01854222185081906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,7168,0.008998221821255153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,256,32,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,7168,0.017252445220947266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,6144,0.008303110798199972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,5120,0.008225777910815345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,6144,0.01625866691271464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,4096,0.0073315559162033936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,5120,0.015465777781274585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,3584,0.006599999964237213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,4096,0.014089778065681458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,3072,0.005962666538026597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,3584,0.013806222213639153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,2560,0.00527022240890397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,3072,0.01313244468635983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,2048,0.004939555707905027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,2560,0.01315111087428199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,1536,0.0041982221106688184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,2048,0.0123831108212471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,1024,0.003924444317817688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,1536,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,768,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,1024,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,512,0.0033475555893447665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,768,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,256,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,512,0.011610666910807291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,128,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,256,0.011358222199810876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,64,0.002534222271707323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,128,0.010716444088353051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,128,32,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,64,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,65536,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,128,32,0.010390222072601318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,51200,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,65536,0.09480177693896824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,16384,0.008993777963850234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,51200,0.07318311267428927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,12288,0.009110222260157267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,16384,0.02845422095722622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,12288,0.023354666100607976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,10240,0.008298666940795051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,10240,0.021278222401936848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,8192,0.00830666638082928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,7168,0.008671110702885522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,8192,0.018512000640233357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,6144,0.008010666403505538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,7168,0.017599110802014668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,6144,0.016576000385814242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,5120,0.007418666448858049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,4096,0.007330666813585494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,5120,0.015202666322390238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,3584,0.0069208890199661255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,4096,0.014299554957283868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,3584,0.01352533366945055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,3072,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,2560,0.006036444256703059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,3072,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,2048,0.005939555664857228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,2560,0.01310755560795466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,2048,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,1536,0.004292444636424382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,1024,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,1536,0.012400888734393649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,768,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,1024,0.011815999945004782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,768,0.01168444421556261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,512,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,256,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,512,0.011519999967681037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,128,0.002798222212327851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,256,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,128,0.01072622256146537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,64,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,64,32,0.0025928889711697898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,64,0.010427555276287926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,65536,0.027124444643656414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,64,32,0.010440000229411654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,51200,0.022032888399230108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,65536,0.09364000293943618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,16384,0.00904622260067198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,51200,0.07258222500483195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,16384,0.02815022071202596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,12288,0.00907199995385276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,10240,0.008415111237102086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,12288,0.023461333579487268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,8192,0.007687110867765214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,10240,0.021746666895018682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,8192,0.018612444400787354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,7168,0.008397333323955536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,6144,0.008006222546100616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,7168,0.017309332887331646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,5120,0.007326222128338284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,6144,0.01660977800687154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,4096,0.006664000037643645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,5120,0.015192889504962497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,3584,0.0064737776087390045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,4096,0.01420088940196567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,3584,0.01347555551264021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,3072,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,2560,0.005960888746711943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,3072,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,2048,0.005633777628342311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,2560,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,1536,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,2048,0.012390221986505719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,1024,0.003921777837806278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,1536,0.012144889268610211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,1024,0.011728000309732227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,768,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,512,0.003245333416594399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,512,0.011591110792424945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,256,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,256,0.011006222003036074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,128,0.002805333377586471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,128,0.01073688848151101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,64,0.002536000063021978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,768,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,129,32,32,0.002611555573013094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,64,0.010445333189434476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,129,32,32,0.010391111175219217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,51200,0.6443849139743382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,51200,1.0778204600016277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,16384,0.36492088105943465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,16384,0.21482488844129774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,12288,0.27126044697231716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,12288,0.16511200533972845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,16384,0.27021778954399955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,10240,0.23173954751756456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,12288,0.2108551131354438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,10240,0.14110399617089167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,8192,0.19188800123002794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,8192,0.11163200272454156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,10240,0.1808506647745768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,7168,0.1620053317811754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,7168,0.10057422187593247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,8192,0.1488693290286594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,6144,0.14374844233194986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,6144,0.08810310893588597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,51200,0.7945129076639811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,5120,0.11832711431715225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,7168,0.13420355319976807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,5120,0.07587110996246338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,6144,0.11853333314259847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,4096,0.09720266527599758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,4096,0.06294933292600843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,3584,0.08502755562464397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,5120,0.10521599981519912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,3584,0.05677600039376152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,4096,0.08954133590062459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,3072,0.07360266976886325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,3072,0.051501333713531494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,2560,0.06286222404903836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,3584,0.08254310819837782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,3072,0.07495288716422187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,2560,0.045259555180867515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,2048,0.05215644505288866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,2560,0.06811022096210055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,2048,0.03977866636382209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,1536,0.03943555553754171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,2048,0.06025511026382446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,1024,0.027796443965699937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,1536,0.03333866596221924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,1536,0.05406577719582451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,768,0.02237333357334137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,1024,0.027222222752041284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,1024,0.04646133383115133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,512,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,768,0.023250665929582384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,768,0.04174044397142199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,256,0.010794666906197866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,512,0.019906666543748643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,512,0.03819288810094198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,128,0.00833422193924586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,256,0.017975999249352347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,256,0.033356444703208074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,64,0.007611555357774098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,128,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,65536,128,0.02664088871743944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,65536,32,0.009733333355850643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,64,0.017154667112562392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,65536,32,0.022364444202846948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,65536,0.6628062460157607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,65536,1.070574230617947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,51200,0.8440044191148547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,51200,0.5164017677307129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,16384,0.2874071068233914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,16384,0.17387289471096465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,16384,0.23625066545274523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,12288,0.21044088734520805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,12288,0.13464266724056667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,65536,0.8702026473151313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,10240,0.176602668232388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,51200,0.686757352617052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,12288,0.18513688776228163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,8192,0.14628622266981336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,10240,0.11509422461191814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,8192,0.09153955512576634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,7168,0.12356444199879964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,10240,0.15843733151753744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,8192,0.13154933187696669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,7168,0.08256355259153578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,6144,0.10945510864257812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,6144,0.07272355424033271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,5120,0.09038044346703424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,7168,0.11786933739980061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,6144,0.1050293313132392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,4096,0.07456444369422065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,5120,0.06302933560477363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,4096,0.052541332112418286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,5120,0.09288089142905341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,3584,0.06557244724697538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,4096,0.08035466406080458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,3584,0.04764622118737963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,3072,0.05707910988065931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,3584,0.07380711370044284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,2560,0.04907822277810839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,3072,0.06790755854712592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,2560,0.03862577676773071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,2048,0.04075110952059428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,2560,0.06203911039564344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,2048,0.03436177637841966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,1536,0.029694222741656836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,2048,0.05495199892255995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,1536,0.029241777128643457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,1024,0.021799110703998145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,1536,0.049511998891830444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,3072,0.04359289010365804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,1024,0.022562666071785822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,768,0.016606221596399944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,1024,0.04270133376121521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,512,0.012176888684431711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,768,0.0199937770764033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,768,0.03903288973702325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,256,0.008728000024954477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,512,0.035006221797731184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,512,0.018232888645595975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,128,0.006284444282452266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,256,0.030304001437293157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,128,0.015855110353893705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,256,0.01648533344268799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,51200,128,0.025248888466093276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,64,0.005631999837027655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,51200,32,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,64,0.017097777790493436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,51200,32,0.01830755505296919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,65536,0.3224853409661187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,65536,0.271688010957506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,51200,0.24638756116231283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,51200,0.21512356069352892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,16384,0.08958666854434544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,16384,0.07678666379716662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,16384,0.08796266714731853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,12288,0.06418666574690077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,65536,0.2751084433661567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,51200,0.21737333138783774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,12288,0.06159822146097819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,10240,0.05406755540106031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,10240,0.053631110323799976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,8192,0.04520622226927015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,12288,0.07247555255889893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,8192,0.04267911116282145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,10240,0.06440089146296184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,7168,0.03992444607946608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,8192,0.0547013349003262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,6144,0.035178667969173856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,7168,0.050085332658555776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,7168,0.039460443788104586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,6144,0.03528533379236857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,5120,0.02988977895842658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,5120,0.03130755490726895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,4096,0.024910221497217815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,5120,0.042747557163238525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,4096,0.027143999934196472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,3584,0.022494221727053326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,3584,0.02402666707833608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,4096,0.03864088985655043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,3072,0.01978488928741879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,3584,0.03614044520590041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,3072,0.034513778156704374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,2560,0.016951110627916124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,6144,0.04624977707862854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,2560,0.020744888318909537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,2560,0.032660444577534996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,2048,0.013956444131003486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,2048,0.030701332622104224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,2048,0.01917866700225406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,1536,0.011556444068749746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,1536,0.017469333277808297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,1536,0.02903111113442315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,1024,0.008750222623348236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,1024,0.01550222271018558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,1024,0.0282008879714542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,768,0.007848888635635376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,3072,0.022846221923828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,512,0.0063822223908371395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,768,0.025111999776628282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,512,0.023051554958025616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,256,0.0042160000238153665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,512,0.013281777501106262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,256,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,128,0.0039822223285834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,256,0.022169777088695105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,128,0.012266666524940066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,64,0.0035911111368073356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,16384,128,0.01944266590807173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,768,0.014186667071448432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,16384,32,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,32,0.01371911085314221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,65536,0.24252533912658691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,65536,0.2207875516679552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,51200,0.18825156158871123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,16384,64,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,51200,0.17556622293260363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,16384,0.06978488630718656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,16384,0.06372888882954915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,65536,0.24182044135199654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,12288,0.04981777734226651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,16384,0.0793822209040324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,51200,0.1934951146443685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,10240,0.04288888971010844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,12288,0.05172266562779745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,12288,0.06554577747980754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,8192,0.035375111632876925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,10240,0.04563111066818237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,10240,0.05919466416041056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,8192,0.03651288814014859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,7168,0.03106044398413764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,8192,0.05049688948525322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,6144,0.027453333139419556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,7168,0.03363822235001458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,7168,0.0463120010164049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,5120,0.024131556351979572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,6144,0.04259377717971802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,5120,0.03992799917856852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,4096,0.019460444649060566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,5120,0.027207111318906147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,4096,0.022960889670583937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,4096,0.03683377636803521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,3584,0.01761866609255473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,6144,0.02986488739649455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,3584,0.021536888347731695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,3072,0.015378667248619927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,3584,0.034726222356160484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,3072,0.033526221911112465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,3072,0.020337777005301583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,2560,0.013212444053755866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,2560,0.032157331705093384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,2048,0.01129066695769628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,2048,0.017803554733594257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,2048,0.030223998758527968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,1536,0.010069333016872406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,1536,0.016551999582184684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,1536,0.028167999453014795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,1024,0.007669332954618666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,2560,0.018933332628673978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,1024,0.014482667048772177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,768,0.006680000159475539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,1024,0.025184889634450276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,768,0.013346666263209449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,512,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,768,0.02474222249454922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,512,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,256,0.004303110970391167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,512,0.02313422163327535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,256,0.012112889024946423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,128,0.003934222377008862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,256,0.021746666895018682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,128,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,64,0.003622222277853224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,12288,128,0.019342222147517733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,12288,32,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,64,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,12288,32,0.01219911128282547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,65536,0.20397510793473986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,65536,0.21410577827029756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,51200,0.15794755352867976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,51200,0.17022577921549478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,65536,0.23139111200968424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,16384,0.0584133333630032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,16384,0.061569776799943715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,16384,0.07632088661193848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,12288,0.042176001601748996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,51200,0.18604800436231825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,12288,0.049866666396458946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,10240,0.03622400098376804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,12288,0.06341689162784152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,10240,0.04440266556209988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,8192,0.02983022067281935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,10240,0.057209776507483594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,8192,0.03511733478969998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,7168,0.02646044393380483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,8192,0.048919998937182956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,7168,0.032300445768568255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,6144,0.023736000061035156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,7168,0.04486755530039469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,6144,0.029410666889614526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,5120,0.020729778541458976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,6144,0.04127466678619385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,5120,0.026294221480687458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,4096,0.016605334149466622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,4096,0.02235200007756551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,5120,0.03864444295565287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,3584,0.014935111006100973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,4096,0.036031110419167414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,3584,0.03395911057790121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,3072,0.013012444807423485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,3584,0.021107556091414556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,3072,0.020069332586394418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,2560,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,3072,0.032708442873424955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,2560,0.018637332651350234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,2048,0.010142222046852112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,2560,0.031319111585617065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,2048,0.01720977822939555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,1536,0.008729777402347988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,2048,0.028928889168633357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,1536,0.015454222758611044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,1024,0.007071110937330458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,1536,0.028798222541809082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,1024,0.01350311107105679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,768,0.00564533347884814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,1024,0.02407466702991062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,768,0.023395554886923894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,512,0.004276444514592488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,512,0.012533333566453723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,512,0.023136888941129048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,256,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,768,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,256,0.020960888928837244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,256,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,128,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,10240,128,0.01888177792231242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,64,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,128,0.011535111400816174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,10240,32,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,64,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,10240,32,0.011832888755533429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,65536,0.1862479978137546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,65536,0.1715395583046807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,51200,0.14509333504570857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,65536,0.21068267027537027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,51200,0.13788621955447727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,16384,0.04866310954093933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,16384,0.05186666713820564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,16384,0.07015555434756808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,51200,0.16966399881574842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,12288,0.041893333196640015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,10240,0.029808001385794744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,12288,0.05962577793333265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,10240,0.03679555654525757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,8192,0.02511111067401038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,10240,0.05392355720202128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,8192,0.029488888051774766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,7168,0.02239911092652215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,8192,0.04656889041264852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,7168,0.027505778604083594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,12288,0.03477688961558872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,6144,0.0199626667631997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,7168,0.04289422101444668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,5120,0.01720888912677765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,6144,0.0240000006225374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,6144,0.040103111002180315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,4096,0.013939556148317126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,5120,0.021962665849261816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,5120,0.037329776419533625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,3584,0.012387555506494312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,4096,0.01994044416480594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,4096,0.03518666823705038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,3072,0.011197333534558615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,3584,0.01917866700225406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,3584,0.03354310989379883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,2560,0.009961778091059791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,3072,0.03234577841228909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,3072,0.018553778529167175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,2048,0.009135111338562435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,2560,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,2560,0.03087555699878269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,1536,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,2048,0.029149333635965984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,2048,0.015860444969601102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,1024,0.005941333456171884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,1536,0.014112888111008538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,1536,0.027048889133665297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,768,0.004920000003443824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,1024,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,1024,0.024350222614076402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,512,0.004182222402758068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,768,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,768,0.023669333921538457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,256,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,512,0.022686221533351477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,512,0.012396444049146442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,128,0.0036355555057525635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,256,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,256,0.02037333283159468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,64,0.003547555456558863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,128,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,8192,32,0.0035955554081334006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,8192,128,0.019881778293185763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,64,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,8192,32,0.01203644441233741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,65536,0.16620622740851507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,51200,0.12788532839881048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,65536,0.16660355197058782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,51200,0.13201155927446154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,16384,0.04176444477505154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,16384,0.05049511128001743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,65536,0.2082071171866523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,51200,0.16780532730950248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,16384,0.06917777988645765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,12288,0.04052266809675429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,10240,0.027354665928416785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,12288,0.05899288919236925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,10240,0.035319109757741295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,8192,0.02253955602645874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,10240,0.053364445765813194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,8192,0.028926223516464233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,7168,0.020071999894248117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,8192,0.04640444450908237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,7168,0.02682311170630985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,12288,0.030910223722457886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,7168,0.04240444302558899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,6144,0.023918223049905565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,6144,0.03989511066012912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,5120,0.015150222513410779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,5120,0.021846221552954778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,5120,0.037109331952200994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,4096,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,4096,0.019708444674809773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,4096,0.03472977876663208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,3584,0.011507555842399597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,3584,0.01871200071440803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,3584,0.033203555477990046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,3072,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,6144,0.017589333984586928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,3072,0.018040888839297824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,3072,0.031940443648232356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,2560,0.016861332787407767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,2560,0.0311119986904992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,2048,0.008252444366614023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,2048,0.01552177800072564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,2048,0.02828622195455763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,1536,0.0069208890199661255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,1536,0.014039110806253223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,2560,0.009368889033794403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,1024,0.0052151112920708126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,1536,0.025098666548728943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,1024,0.02444533341460758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,1024,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,768,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,768,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,768,0.024358222881952923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,512,0.003924444317817688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,512,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,512,0.02276444435119629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,256,0.0034906665484110513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,256,0.011701333026091257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,256,0.021330666210916307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,128,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,7168,128,0.01831999917825063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,64,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,64,0.011703111231327057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,7168,32,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,32,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,65536,0.14507555961608887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,65536,0.16129333443111843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,7168,128,0.011502222054534487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,51200,0.11282755268944634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,51200,0.12751021650102404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,65536,0.20637422137790254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,16384,0.03887733485963609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,51200,0.16586221588982478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,16384,0.06835555367999606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,12288,0.028049776951471966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,12288,0.03895555602179633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,10240,0.02476355599032508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,12288,0.057578669653998486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,16384,0.04884533418549431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,10240,0.05238577723503113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,8192,0.02068177858988444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,8192,0.027287110686302185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,8192,0.04582133226924472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,7168,0.018237334158685472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,7168,0.026208889153268602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,7168,0.04168444540765551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,6144,0.01611199975013733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,6144,0.023390221926901076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,5120,0.0140666663646698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,6144,0.03938222262594435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,5120,0.02165599995189243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,5120,0.03680800067053901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,10240,0.03327644533581204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,4096,0.019466666711701285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,3584,0.011153777440388998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,3584,0.018269333574506972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,4096,0.03403288788265652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,3584,0.0328800015979343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,3072,0.009611555271678502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,3072,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,2560,0.009089777866999308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,4096,0.01166488892502255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,3072,0.03181155522664388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,2560,0.016585777203241985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,2048,0.007680888805124495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,2560,0.030211554633246526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,1536,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,2048,0.014525334040323893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,2048,0.026047110557556152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,1536,0.013947554760509066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,1536,0.02509333358870612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,1024,0.004790222065316306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,768,0.004407999830113517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,1024,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,1024,0.024088889360427856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,512,0.0038399998512532976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,768,0.02443466583887736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,768,0.012427555190192329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,512,0.012158222496509552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,256,0.0036319999231232535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,512,0.02371911042266422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,128,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,256,0.01145866678820716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,256,0.021529777182473078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,64,0.003378666730390655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,64,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,6144,32,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,32,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,65536,0.11706577406989203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,6144,128,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,65536,0.15858488612704807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,51200,0.09287555350197686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,51200,0.1273368861940172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,65536,0.20392266909281412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,16384,0.038176000118255615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,16384,0.0480693346924252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,6144,128,0.01861422260602315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,12288,0.030790223015679255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,51200,0.16421511438157824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,16384,0.06788622008429633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,12288,0.03718577822049459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,12288,0.057327999009026416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,10240,0.03226844469706217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,10240,0.051423999998304576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,8192,0.01883555617597368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,8192,0.02680355476008521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,10240,0.022669333550665114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,8192,0.04494488901562161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,7168,0.01680355601840549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,7168,0.04122133387459649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,6144,0.014565333724021912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,6144,0.023183999790085688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,6144,0.03862311111556159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,5120,0.013240888714790344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,5120,0.02113155523935954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,5120,0.03678933448261685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,4096,0.010739555789364709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,7168,0.025079111258188885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,4096,0.01920088960064782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,3584,0.010044444766309526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,4096,0.033597333563698664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,3584,0.01767200065983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,3584,0.03289599882231818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,3072,0.009200000100665623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,3072,0.031007111072540283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,3072,0.01718311177359687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,2560,0.008442666795518663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,2560,0.02819377846188015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,2560,0.014941333068741692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,2048,0.007666666474607255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,2048,0.027871999475691054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,2048,0.014127999544143677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,1536,0.006670222100284364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,1536,0.02514666650030348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,1536,0.013467555244763693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,1024,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,1024,0.023795555035273235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,768,0.0042160000238153665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,768,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,768,0.022913777165942725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,512,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,512,0.022414222359657288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,256,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,1024,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,256,0.011542222566074796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,128,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,256,0.020994666549894545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,128,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,64,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,512,0.004053333153327306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,5120,32,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,5120,128,0.018912000788582694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,64,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,65536,0.10288088851504856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,5120,32,0.012259555359681448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,65536,0.1188888947168986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,51200,0.07791199949052599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,51200,0.0963475571738349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,16384,0.0322675539387597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,65536,0.20254400041368273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,51200,0.16456354988945857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,12288,0.025920889443821375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,16384,0.06719466712739733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,12288,0.02886755598915948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,12288,0.05693066782421536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,10240,0.019355555375417072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,10240,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,10240,0.051775111092461475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,8192,0.016005333926942613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,8192,0.04432177874777052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,8192,0.021853332718213398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,7168,0.013838222457302941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,7168,0.04145688811937968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,7168,0.020582222276263766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,6144,0.0125608891248703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,6144,0.019280888968043856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,5120,0.011415111521879831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,5120,0.01797066628932953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,16384,0.03600800037384033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,4096,0.00961866643693712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,5120,0.036524444818496704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,4096,0.01627288924323188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,3584,0.009229333036475712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,4096,0.034064001507229276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,3584,0.015336889359686108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,6144,0.03850222296184964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,3584,0.032687111033333674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,3072,0.014517333772447375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,3072,0.029614223374260798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,2560,0.007708444363541073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,2560,0.014633778068754407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,2560,0.02988977895842658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,2048,0.006680888848172293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,2048,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,2048,0.025916443930731878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,3072,0.008392888638708327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,1536,0.012848888834317526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,1536,0.024346666203604803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,1024,0.012157333393891653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,1024,0.023736889163653057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,1536,0.0053182223604785065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,768,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,1024,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,768,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,512,0.003650666524966558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,768,0.0236488895283805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,512,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,256,0.003628444340493944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,512,0.02307022280163235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,128,0.0036888888312710654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,256,0.02233244478702545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,256,0.011101333631409539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,64,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,4096,128,0.019304000669055514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,128,0.011468444433477191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,4096,32,0.0033857776886887024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,64,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,65536,0.09072888559765285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,4096,32,0.011557333171367645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,51200,0.06974133518007067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,65536,0.11612978246476914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,51200,0.09408799807230632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,65536,0.2021991146935357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,16384,0.026896889011065166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,51200,0.16309155358208552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,16384,0.06722578075197008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,12288,0.022074666288163926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,16384,0.03494844502872891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,10240,0.019428445233239066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,12288,0.02827289038234287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,12288,0.05646577808592054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,8192,0.014831110835075378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,10240,0.02545333405335744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,10240,0.05110399921735128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,7168,0.013291555146376291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,8192,0.02150933278931512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,8192,0.04470755656560262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,6144,0.012324444121784635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,7168,0.020389333367347717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,7168,0.04090222385194566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,5120,0.010647999743620554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,6144,0.03872977693875631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,6144,0.018960000740157235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,4096,0.009381333159075843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,5120,0.03603911068704393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,5120,0.01774666706720988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,3584,0.008408000071843466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,4096,0.015855999456511605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,4096,0.034119112624062434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,3072,0.007709333466158972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,3584,0.015542222393883599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,3584,0.03271911210483975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,2560,0.007163555257850223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,3072,0.014509333504570855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,3072,0.03000266684426202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,2560,0.02781244450145298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,2048,0.006036444256703059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,2560,0.013888888888888888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,1536,0.005251555393139522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,2048,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,2048,0.02660977840423584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,1024,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,1536,0.012815111213260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,1536,0.02479022244612376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,768,0.003941333128346337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,1024,0.012135110795497894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,1024,0.02365600069363912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,512,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,768,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,768,0.022749332918061152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,512,0.02327466673321194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,256,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,512,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,128,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,256,0.011494222614500256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,256,0.02126755482620663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,64,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3584,128,0.019343111250135634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3584,32,0.0033484444850020935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,64,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,32,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,65536,0.0729973316192627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,65536,0.11258044507768418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,51200,0.0598906675974528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,65536,0.20195466942257354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,51200,0.09142488903469509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,16384,0.025771554973390367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3584,128,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,16384,0.03334310981962416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,12288,0.020573332905769348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,51200,0.16309421592288548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,16384,0.06711200210783216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,10240,0.01779111060831282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,12288,0.027436445156733196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,12288,0.055868446826934814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,8192,0.01407999959256914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,10240,0.025137777129809063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,10240,0.05092533429463705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,7168,0.012835555606418185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,8192,0.02109777761830224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,8192,0.044601778189341225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,7168,0.01997599999109904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,6144,0.011527111132939657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,7168,0.04115110966894362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,6144,0.018798222144444782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,5120,0.010508444574144153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,6144,0.038853334056006536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,5120,0.017281777328915067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,4096,0.008943999807039896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,5120,0.03630222214592828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,4096,0.01517777807182736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,3584,0.008548444343937768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,4096,0.033328887489106916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,3584,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,3072,0.007947555846638149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,3584,0.031903111272388034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,3072,0.01386755539311303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,3072,0.029508444997999404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,2560,0.013468444347381592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,2560,0.027878221538331773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,2048,0.005642666584915585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,2048,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,2560,0.006980444822046492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,2048,0.026196445027987163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,1536,0.004958222309748332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,1536,0.02478844424088796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,1536,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,1024,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,1024,0.023725334140989516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,1024,0.012166221936543783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,768,0.003929777691761653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,768,0.02397777802414364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,768,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,512,0.003972444269392225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,512,0.02293955617480808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,512,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,256,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,256,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,256,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,128,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,3072,128,0.019003555178642273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,128,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,64,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,3072,32,0.003489777859714296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,64,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,3072,32,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,65536,0.06554933389027913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,65536,0.10954044262568156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,51200,0.05199822121196323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,65536,0.2022693289650811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,51200,0.08893333541022407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,16384,0.021951110826598272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,16384,0.03198755449718899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,51200,0.16329422261979845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,12288,0.018611555298169453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,16384,0.06642666790220472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,10240,0.016653334101041157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,12288,0.026943110757403906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,12288,0.05662755833731758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,8192,0.01386577801571952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,10240,0.02440622283352746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,10240,0.05079911152521769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,7168,0.013259555730554791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,8192,0.04416444566514757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,8192,0.020978666014141504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,6144,0.011129777464601727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,7168,0.01960088809331258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,7168,0.04088266690572103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,5120,0.009697777529557547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,6144,0.018212444252438016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,6144,0.03802488909827338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,4096,0.008575110799736446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,5120,0.016757332616382174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,5120,0.035616000493367515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,4096,0.014834667245546976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,3584,0.008207110895050896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,4096,0.03163466519779629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,3072,0.007150222029950883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,3584,0.014087999860445658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,3584,0.030272887812720403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,3072,0.01384622189733717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,2560,0.006688888702127669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,3072,0.028504000769721136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,2048,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,2560,0.013447999954223633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,2560,0.02850311001141866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,1536,0.004943999979231092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,2048,0.012818666795889536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,2048,0.026143999563323125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,1024,0.004409777621428172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,1536,0.024353777368863423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,1536,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,1024,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,768,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,1024,0.023503111468421087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,512,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,768,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,768,0.023355555203225877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,256,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,512,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,512,0.023119111855824787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,256,0.01148888882663515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,256,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,128,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,128,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2560,128,0.019280888968043856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,64,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2560,32,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,64,0.011388444238238864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2560,32,0.011158222125636207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,65536,0.05697155661053128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,65536,0.10649777783287896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,51200,0.044826666514078774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,65536,0.20179643895890978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,51200,0.08619644244511922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,16384,0.01833777791923947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,16384,0.03135910961363051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,16384,0.0664551125632392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,12288,0.014612444572978549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,12288,0.026836444934209187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,12288,0.05583644575542874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,10240,0.013910222384664746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,51200,0.16325688362121582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,10240,0.02442488902144962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,8192,0.013261333107948303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,10240,0.05130399929152595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,8192,0.02054755555258857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,7168,0.011993777420785693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,8192,0.04439022143681844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,7168,0.018957333432303537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,6144,0.010778666370444827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,7168,0.040756444136301674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,6144,0.017606221967273288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,5120,0.009714666340086196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,6144,0.03735111157099406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,5120,0.01587466730011834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,4096,0.008400888906584846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,5120,0.0343484448062049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,4096,0.014557333456145393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,4096,0.031609776947233416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,3584,0.007642666498819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,3584,0.029616889026429918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,3584,0.014200000299347771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,3072,0.0069466663731469055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,3072,0.02885066800647312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,3072,0.01387288918097814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,2560,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,2560,0.027237332529491846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,2560,0.01348977784315745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,2048,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,2048,0.026093333959579468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,2048,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,1536,0.004657777647177379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,1536,0.02473066747188568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,1536,0.012740444805887012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,1024,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,1024,0.023976888921525743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,1024,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,768,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,768,0.023775999744733173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,768,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,512,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,512,0.023149333066410486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,512,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,256,0.020640888147883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,128,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,128,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,2048,128,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,64,0.0028435554769304064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,64,0.011111111276679568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,2048,32,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,32,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,2048,256,0.011415111521879831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,65536,0.042691555288102895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,65536,0.10423288742701213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,65536,0.20195999410417345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,51200,0.03449777762095133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,51200,0.0838551123936971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,16384,0.015494222442309061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,51200,0.16369333532121447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,16384,0.031172444423039753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,16384,0.06672533353169759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,12288,0.012686221963829465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,12288,0.055626663896772593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,12288,0.026533333791626826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,10240,0.011868444581826528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,10240,0.050612443023257785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,8192,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,8192,0.01960533360640208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,8192,0.04424800144301521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,7168,0.010068444742096795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,7168,0.018719111879666645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,7168,0.03981688949796889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,10240,0.024000888069470722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,6144,0.010716444088353051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,6144,0.03639555639690823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,5120,0.009330666727489894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,5120,0.015858666764365304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,5120,0.03369777732425266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,6144,0.017218665944205392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,4096,0.008005333443482717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,4096,0.0148026661740409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,4096,0.031215111414591473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,3584,0.006999111009968652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,3584,0.014127110441525778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,3584,0.030213332838482324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,3072,0.006604444649484422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,3072,0.013840888937314352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,2560,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,3072,0.028679112593332928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,2560,0.01350488927629259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,2048,0.005400889035728242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,2560,0.02720533311367035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,2048,0.026149332523345947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,2048,0.012914666699038612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,1536,0.004874666531880696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,1536,0.02409422232045068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,1536,0.012730666332774691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,1024,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,1024,0.02367288867632548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,1024,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,768,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,768,0.02276800076166789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,768,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,512,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,512,0.02202844454182519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,512,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,256,0.0033164444482988785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,256,0.020015999674797058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,256,0.01144177797767851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,128,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1536,128,0.01827911039193471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,128,0.011343110766675739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,64,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1536,32,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,64,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1536,32,0.01076533314254549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,65536,0.031182222896152075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,65536,0.10067199998431736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,51200,0.025824889540672302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,65536,0.20206577248043486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,51200,0.08123022317886353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,51200,0.1636293331782023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,16384,0.012112889024946423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,16384,0.06644088692135282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,16384,0.030957334571414526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,12288,0.025340444511837427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,12288,0.056206219726138644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,10240,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,10240,0.023120888405376013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,12288,0.013090666797426013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,8192,0.01093066649304496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,8192,0.018965333700180054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,8192,0.04196977615356445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,7168,0.010021333065297868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,7168,0.018030222919252183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,7168,0.03824444280730353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,10240,0.050717333952585854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,6144,0.00903733323017756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,6144,0.03604799840185378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,6144,0.017080000705189176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,5120,0.008016889293988546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,5120,0.03366400135887994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,5120,0.015810666812790763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,4096,0.0074311114019817775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,4096,0.014891554911931356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,4096,0.031023999055226643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,3584,0.00700711127784517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,3584,0.013908444179428948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,3584,0.03018933203485277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,3072,0.013793778088357715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,3072,0.028951111767027114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,2560,0.00565422202150027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,2560,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,3072,0.006393777827421824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,2560,0.02815999918513828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,2048,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,2048,0.025835555460717943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,2048,0.012810666528013019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,1536,0.004952888935804367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,1536,0.024127999941507976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,1536,0.012741333080662621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,1024,0.0042408886882993905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,1024,0.023416888382699754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,1024,0.012055111428101858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,768,0.00388355553150177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,768,0.022737777895397607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,768,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,512,0.0034515555534097883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,512,0.021963554951879714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,512,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,256,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,256,0.0199955552816391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,256,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,1024,128,0.018280888597170513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,128,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,32,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,64,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,1024,32,0.010783111055692038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,65536,0.026485333840052288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,65536,0.09978222184711033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,51200,0.02144444485505422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,1024,64,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,65536,0.20216266314188638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,51200,0.07998844650056627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,51200,0.16319555706448025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,16384,0.012867555850081973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,12288,0.011182222101423474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,16384,0.0658408866988288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,12288,0.024178667200936213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,12288,0.054735110865698926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,10240,0.02203377750184801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,10240,0.048284444544050426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,8192,0.008968888885445064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,8192,0.018900444110234577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,16384,0.030177778667873804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,7168,0.008565333154466417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,10240,0.010108444425794814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,7168,0.038689778910742864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,6144,0.007603555917739868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,7168,0.017966222431924608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,6144,0.01686844395266639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,5120,0.008280889027648503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,6144,0.03573244478967454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,5120,0.015818667080667283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,4096,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,8192,0.04261689053641426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,5120,0.033583998680114746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,4096,0.014526221487257214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,3584,0.007122666471534305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,4096,0.030564443932639226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,3072,0.00638133328821924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,3584,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,3584,0.02946399980121189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,2560,0.005759999983840519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,3072,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,3072,0.028507557180192735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,2048,0.00518222235971027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,2560,0.027331555883089702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,2560,0.013188444077968597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,1536,0.004634666774008009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,2048,0.012736000120639801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,2048,0.025755556093321905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,1024,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,1536,0.02478222217824724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,1536,0.012288889123333825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,768,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,1024,0.011775111158688864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,1024,0.02372088862790002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,512,0.003367111086845398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,768,0.02235200007756551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,768,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,256,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,512,0.011712888876597086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,512,0.02200177808602651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,128,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,256,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,256,0.021319111188252766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,64,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,128,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,768,32,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,768,128,0.018568888306617737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,64,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,65536,0.019001776973406475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,768,32,0.010776888993051318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,51200,0.015801777442296345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,65536,0.0964168906211853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,65536,0.2018524408340454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,16384,0.010445333189434476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,51200,0.07822933461931017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,51200,0.16288799709743923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,12288,0.009214222431182861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,16384,0.0649520026312934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,16384,0.02880355715751648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,10240,0.009310222334331935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,12288,0.024047111471494038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,12288,0.05311110946867201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,8192,0.008316444853941599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,10240,0.02162933349609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,10240,0.04784711201985677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,7168,0.007620444728268518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,8192,0.01863822175396813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,8192,0.04140977727042304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,6144,0.007307555940416124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,7168,0.037738667594061956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,7168,0.017845332622528076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,5120,0.007648888561460707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,6144,0.016754666964213055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,6144,0.03569688730769687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,4096,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,5120,0.015825778245925903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,5120,0.033258666594823204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,3584,0.0069475554757648045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,4096,0.014136888914638095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,4096,0.031179553932613794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,3072,0.006305777778228124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,3584,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,3584,0.02955733405219184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,3072,0.013422222601042854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,2560,0.005568888866239124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,3072,0.028466665082507666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,2560,0.013095999757448832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,2048,0.004939555707905027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,2560,0.027241778042581346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,2048,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,1536,0.004583111239804162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,2048,0.025477333201302424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,1536,0.012409778104888068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,1024,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,1536,0.024775111012988623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,1024,0.012048888537618848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,768,0.003549333247873518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,1024,0.022708444131745234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,768,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,512,0.003504000190231535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,512,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,768,0.022454222043355305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,256,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,512,0.023090667194790308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,256,0.021007999777793884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,128,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,256,0.011194667054547204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,128,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,64,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,512,128,0.018738667170206707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,512,32,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,64,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,65536,0.01519733336236742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,512,32,0.010423111418883005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,65536,0.09619733360078599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,51200,0.013210666676362356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,65536,0.20188621679941812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,51200,0.16218311256832546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,16384,0.009351111120647853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,51200,0.07531644238366021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,16384,0.028412444723976985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,12288,0.008460444708665213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,16384,0.06334844562742445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,10240,0.0087333329849773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,12288,0.023881778120994568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,12288,0.05382755398750305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,8192,0.008042666647169325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,10240,0.047664887375301786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,10240,0.02160444524553087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,7168,0.0075902218619982404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,8192,0.018996444013383653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,8192,0.04216622312863668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,6144,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,7168,0.01759999990463257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,7168,0.03872355487611558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,6144,0.01663466625743442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,5120,0.007739555504586961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,6144,0.035375111632876925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,5120,0.015771556231710646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,4096,0.007320000065697565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,5120,0.03394044438997904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,3584,0.007041778001520369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,4096,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,4096,0.030538668235143025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,3072,0.006341333190600078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,3584,0.014161777165200977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,3584,0.029912001556820337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,2560,0.006127999888526068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,3072,0.013434666726324292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,3072,0.02892355455292596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,2560,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,2048,0.005991999887757831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,2560,0.027094221777386133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,2048,0.012793777717484368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,1536,0.004583111239804162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,2048,0.026622222529517278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,1024,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,1536,0.012401777837011548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,1536,0.02513422237502204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,768,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,1024,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,1024,0.023808000816239253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,512,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,768,0.023335110810067918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,768,0.011448889142937131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,512,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,256,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,512,0.022916444473796423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,256,0.01146666705608368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,128,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,256,0.02087822225358751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,64,0.0025395556456512874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,128,0.011133333047231039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,256,128,0.0196088883611891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,256,32,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,65536,0.014782221780882942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,64,0.010557333628336588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,256,32,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,51200,0.014527110589875115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,65536,0.09240000115500556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,65536,0.20155910650889078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,16384,0.009016888837019602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,51200,0.0745297802819146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,51200,0.16140088770124647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,16384,0.06369511286417644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,16384,0.028262221150928076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,12288,0.007985778152942657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,12288,0.05308533377117581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,12288,0.023919999599456787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,10240,0.007328000333574083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,10240,0.04801955487993029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,8192,0.006993778049945831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,10240,0.021709332863489788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,8192,0.018938667244381376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,7168,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,8192,0.041872001356548734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,7168,0.017627555463049147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,6144,0.006330666856633292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,7168,0.03820711043145921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,6144,0.016536000702116225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,5120,0.0063324446479479475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,5120,0.015583111180199517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,6144,0.03580266568395827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,4096,0.006436444405052397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,5120,0.03360355562633938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,4096,0.03092355529467265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,3584,0.00639111093348927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,4096,0.01424800025092231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,3584,0.013879111243618859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,3072,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,3584,0.029701332251230877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,3072,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,2560,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,3072,0.028612444798151653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,2560,0.013175999952687157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,2048,0.005994666781690385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,2560,0.027275555663638647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,2048,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,1536,0.004486222234037188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,2048,0.025288888149791296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,1536,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,1024,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,1536,0.025125333004527624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,1024,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,768,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,1024,0.022839110758569505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,768,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,512,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,512,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,768,0.02288444505797492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,256,0.002995555599530538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,512,0.022870222727457683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,256,0.021301334102948506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,256,0.010873777998818291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,128,0.0028719999310043124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,128,128,128,0.018944000204404194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,64,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,128,0.010773333410422007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,128,32,0.002704888789190186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,65536,0.013576889203654395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,64,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,51200,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,128,32,0.010424888796276517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,65536,0.0940284464094374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,51200,0.073125335905287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,16384,0.007848000360859765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,12288,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,10240,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,16384,0.028113779094484117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,8192,0.006919999917348226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,10240,0.021663111117151048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,7168,0.006609777609507243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,8192,0.018252443936136033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,12288,0.02388622197839949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,6144,0.006276444428496891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,5120,0.005966222120655908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,7168,0.01759022143152025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,4096,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,5120,0.015120888749758402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,6144,0.016538666354285348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,3584,0.006252444452709622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,4096,0.014506666196717156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,3072,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,2560,0.005956444475385878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,3072,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,2048,0.005888000130653381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,2560,0.012805333567990197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,3584,0.01368533323208491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,1536,0.004268444246715969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,1024,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,2048,0.012828444441159567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,768,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,1024,0.011689778003427716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,512,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,1536,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,768,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,256,0.002983999955985281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,512,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,128,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,256,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,64,0.002591111179855135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,64,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,128,0.01069866700304879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,64,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,65536,0.012839111189047495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,64,32,0.010453333457310995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,51200,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,51200,0.07183021969265409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,16384,0.007898666792445714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,16384,0.02830044428507487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,12288,0.007365333537260692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,65536,0.09175822469923232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,10240,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,12288,0.023733332753181458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,10240,0.021721776988771226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,8192,0.006666666517655055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,7168,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,8192,0.01866133345497979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,6144,0.0063777777055899305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,7168,0.01761155492729611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,6144,0.01667644414636824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,5120,0.01587377819750044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,4096,0.005922666854328579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,3584,0.006537777682145436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,3584,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,3072,0.005953777581453323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,5120,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,3072,0.013752000199423896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,2560,0.005707555760939916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,2048,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,2560,0.01348000019788742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,1536,0.00452888881166776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,2048,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,4096,0.014452444182501899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,1024,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,768,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,1536,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,1024,0.012129777835475074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,512,0.0033537778589460584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,256,0.002983999955985281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,768,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,128,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,256,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,64,0.0028026666906144884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,128,0.010458666417333813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,128,32,32,0.002568888788421949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,64,0.011138666835096149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,32,0.010391111175219217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,128,32,512,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,51200,0.638980441623264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,16384,0.3678737746344672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,51200,1.0893270704481337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,12288,0.26947911580403644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,16384,0.21211911572350395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,10240,0.23242934544881186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,12288,0.16362399525112575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,10240,0.13805333773295084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,8192,0.19210133287641737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,8192,0.11085600323147243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,7168,0.1615484423107571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,7168,0.09929422537485759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,6144,0.14333332909478083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,5120,0.11775822109646267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,6144,0.08732710944281684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,5120,0.07519111368391249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,3584,0.08484711249669392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,4096,0.06274222003089057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,3584,0.05678577555550469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,3072,0.07288355297512479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,2560,0.06275377670923869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,3072,0.05134755704138014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,2048,0.05150666501786974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,4096,0.09742044740253025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,2560,0.04481777879926893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,1536,0.03912622067663405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,2048,0.03950577643182542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,1024,0.027839110957251653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,768,0.02197422252760993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,1536,0.03324177861213684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,512,0.01552888916598426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,1024,0.028202666176689997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,768,0.024352888266245525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,256,0.010537777509954242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,128,0.008018666671382057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,512,0.0196088883611891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,256,0.01830222209294637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,64,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,65536,32,0.008126222425036961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,128,0.01717866626050737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,64,0.01720622181892395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,65536,32,0.019650666250122916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,65536,0.6560666826036241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,65536,1.06731383005778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,51200,0.8391946686638726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,16384,0.28641245100233287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,16384,0.1716222233242459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,51200,0.5096311039394802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,12288,0.2118675576315986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,10240,0.175528883934021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,12288,0.13288711176978216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,8192,0.14625955952538386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,7168,0.12311200300852458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,8192,0.09048622184329563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,10240,0.11240977711147732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,6144,0.10929155349731445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,7168,0.08144711123572455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,5120,0.09061600102318658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,6144,0.07209866576724581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,4096,0.07374755541483562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,5120,0.06233421961466471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,3584,0.06542222367392646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,4096,0.05253511004977756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,3584,0.04750400119357639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,3072,0.0565884444448683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,2560,0.04879822333653768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,3072,0.04320089022318522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,2048,0.04073422153790792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,2560,0.038796444733937584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,1536,0.029818667305840388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,1024,0.021800888909233943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,2048,0.03378666771782769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,1536,0.029453333881166246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,768,0.016656888855828177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,1024,0.023245332969559565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,512,0.012240889171759287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,256,0.008762666748629676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,768,0.021722666091389124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,512,0.018176888426144917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,128,0.006709333509206772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,256,0.016187555260128446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,64,0.006258666515350342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,51200,32,0.006265777680608962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,128,0.015824889143308003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,64,0.01688088807794783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,51200,32,0.018252443936136033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,65536,0.3209644423590766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,51200,0.24603290028042266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,65536,0.26650844679938424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,16384,0.0911288857460022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,51200,0.20971288945939806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,16384,0.0754293335808648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,10240,0.054690665668911405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,12288,0.06019466453128391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,8192,0.045048889186647206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,10240,0.05021333363321093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,7168,0.03986400034692552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,8192,0.042066666815016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,6144,0.03526488939921061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,7168,0.038559112283918597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,5120,0.030004445049497817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,6144,0.03480888737572564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,4096,0.024792000651359558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,5120,0.030905776553683813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,3584,0.02240000002914005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,4096,0.026569777064853247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,12288,0.064538664287991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,3072,0.019719110594855416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,2560,0.016948444975747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,3584,0.024707555770874023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,2048,0.013744889034165276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,2560,0.020829333199395072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,3072,0.02295822236273024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,1536,0.011407111254003314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,2048,0.018997333116001554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,1024,0.009026666482289633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,768,0.007692444655630324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,1536,0.017655111021465726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,1024,0.015563555889659457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,512,0.005956444475385878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,256,0.004310222135649787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,512,0.013389333254761167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,128,0.003850666599141227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,256,0.01278222186697854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,64,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,128,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,16384,32,0.003826666623353958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,64,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,32,0.013059555656380124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,65536,0.24392355812920463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,65536,0.21579199367099336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,16384,768,0.014462222655614218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,16384,0.06921866867277357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,51200,0.16977155208587646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,12288,0.04997511042488945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,16384,0.06308888726764254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,10240,0.04270311196645101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,12288,0.05065955718358358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,8192,0.03514488869243198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,10240,0.04228533307711283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,7168,0.031209776798884075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,8192,0.035912000470691256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,7168,0.0329022208849589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,6144,0.02743911080890232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,51200,0.1876915560828315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,5120,0.02372444503837162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,4096,0.019339554839664035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,6144,0.029763556189007227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,3584,0.01755555636352963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,4096,0.02267733381854163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,5120,0.025881777207056682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,3584,0.021324444148275588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,2560,0.01330933305952284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,3072,0.020272889071040683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,2560,0.01897333396805657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,2048,0.011359111302428775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,2048,0.01752444439464145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,1536,0.009666666388511658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,1536,0.016565332810084026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,1024,0.007679111427730984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,1024,0.014231999715169271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,3072,0.015433778365453085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,768,0.006583111153708563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,512,0.0047155556579430895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,768,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,256,0.004263111286693149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,512,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,128,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,256,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,64,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,128,0.012235555383894177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,12288,32,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,64,0.01204355557759603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,12288,32,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,65536,0.20303644074334037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,51200,0.15775911013285318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,65536,0.20829689502716064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,16384,0.058614220884111196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,16384,0.06096800168355306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,51200,0.16484710905287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,12288,0.042019555966059365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,10240,0.03611200054486593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,12288,0.04908444484074911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,8192,0.029823111163245306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,10240,0.0411137772930993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,7168,0.026549332671695288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,8192,0.034671111239327326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,6144,0.023730667101012334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,7168,0.03193599979082743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,5120,0.020470221837361652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,6144,0.027691556347741023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,4096,0.016531555189026725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,5120,0.024677332904603746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,3584,0.014670222997665405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,4096,0.0221742226017846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,3584,0.020969778299331665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,3072,0.013282666603724161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,2560,0.011683555940786997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,3072,0.01973600023322635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,2048,0.00999111102686988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,2560,0.018552889426549275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,1536,0.008496888809733922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,2048,0.01718133356836107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,1024,0.006973333656787872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,1536,0.015477332803938123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,768,0.0052764444715446895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,1024,0.01387644476360745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,512,0.004586666822433472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,768,0.013100444442696042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,256,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,512,0.012825777961148156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,128,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,256,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,64,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,128,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,10240,32,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,64,0.012063110868136087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,10240,32,0.012402666939629449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,65536,0.1809004412757026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,51200,0.13950933350457087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,65536,0.16497155030568442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,16384,0.05133600036303202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,51200,0.13116799460517034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,12288,0.0377422206931644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,16384,0.050252444214291044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,10240,0.03259822063975864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,12288,0.04083999991416931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,8192,0.027800000376171533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,10240,0.03409689002566867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,7168,0.023844444089465674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,8192,0.02863466739654541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,6144,0.021707556313938562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,7168,0.02646844420168135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,5120,0.018259555101394653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,6144,0.02390577726893955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,4096,0.015657777587572735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,5120,0.02174577779240078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,3584,0.012886222038004132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,4096,0.019719110594855416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,3072,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,3584,0.018731556004948087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,2560,0.010281778044170803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,3072,0.017858665850427415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,2048,0.008922666311264038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,1536,0.007709333466158972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,2560,0.01695644524362352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,1024,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,2048,0.01553066737122006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,768,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,1536,0.01383733335468504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,512,0.004271111140648524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,768,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,256,0.0038417776425679526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,512,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,256,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,128,0.0036355555057525635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,64,0.003645333151022593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,64,0.011493333511882357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,8192,32,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,1024,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,8192,32,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,65536,0.16221511363983154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,51200,0.12657333744896784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,65536,0.16284533341725668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,16384,0.042632887760798134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,51200,0.12880799505445692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,16384,0.048932443062464394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,12288,0.031228442986806233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,10240,0.026871111657884385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,12288,0.03970044520166185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,10240,0.032799998919169106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,8192,0.022571555442280237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,7168,0.019919999771647982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,8192,0.028558221128251817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,7168,0.025736888249715168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,6144,0.017511111166742113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,5120,0.015173332558737861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,6144,0.023611555496851604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,5120,0.021369778447681006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,4096,0.012497777740160624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,3584,0.011493333511882357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,4096,0.019620445039537217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,3072,0.010399999717871347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,3584,0.018607111440764535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,2560,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,3072,0.017659554878870647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,2048,0.008199999729792276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,2560,0.016561778055297006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,1536,0.007052444749408298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,2048,0.01517955627706316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,1024,0.005006222261322869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,1536,0.014076444837782117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,768,0.004484444442722532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,1024,0.013134222063753339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,512,0.003931555483076307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,768,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,256,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,512,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,128,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,256,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,64,0.0031555555760860443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,128,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,7168,32,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,64,0.011143999795118967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,7168,32,0.012077333198653327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,65536,0.14248000250922307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,51200,0.11220710807376438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,65536,0.1568791071573893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,16384,0.04397777716318766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,51200,0.12442399395836724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,12288,0.02850488821665446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,16384,0.047648890150917896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,10240,0.024540444215138752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,12288,0.0379688905345069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,8192,0.02049511174360911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,10240,0.032030221488740705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,7168,0.018237334158685472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,8192,0.02684533264901903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,6144,0.016188444362746347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,7168,0.024875554773542616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,5120,0.014228444960382251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,6144,0.023048889305856492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,4096,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,5120,0.021067554752031963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,3584,0.011002666420406766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,4096,0.019270223047998216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,3072,0.009867555565304225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,3584,0.018375999397701688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,2560,0.009011555876996782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,3072,0.017287110288937885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,2048,0.008021333151393467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,2560,0.016010666886965435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,1536,0.0063368889192740125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,2048,0.014154665999942355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,1024,0.004899555610285865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,1536,0.013762666947311826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,768,0.00453599997692638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,1024,0.012818666795889536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,512,0.00423911131090588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,768,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,256,0.003625777860482534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,512,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,128,0.003527111063400904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,256,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,64,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,128,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,6144,32,0.003141333245568805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,64,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,6144,32,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,65536,0.11626577377319336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,51200,0.09201866388320923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,65536,0.15316532717810735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,16384,0.037965334124035306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,51200,0.1223520040512085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,16384,0.04649066593911913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,12288,0.030770666069454614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,12288,0.0362462235821618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,10240,0.02622311148378584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,8192,0.019262222780121695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,10240,0.030621333254708186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,7168,0.016514667206340365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,7168,0.02446577780776554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,6144,0.014817777607176038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,8192,0.02661777701642778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,5120,0.012917333179050021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,6144,0.022667555345429316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,4096,0.010643555886215635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,5120,0.02090666691462199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,3584,0.00996888925631841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,4096,0.018933332628673978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,3072,0.009031111167536842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,3584,0.018248000078731112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,3072,0.016487111647923786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,2560,0.00852177788813909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,2048,0.007328000333574083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,2560,0.015229332778188916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,1536,0.005675555517276128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,2048,0.014532445205582513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,1024,0.004905777672926585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,1536,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,768,0.00425155543618732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,1024,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,512,0.004282666577233208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,768,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,256,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,512,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,256,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,128,0.0035031110876136353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,64,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,128,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,5120,32,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,64,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,5120,32,0.01184977756606208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,65536,0.0991520020696852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,51200,0.07531822390026517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,65536,0.11273688740200466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,16384,0.031832890378104314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,51200,0.0912791093190511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,12288,0.025616889198621113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,16384,0.03450399968359206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,10240,0.02260355485810174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,12288,0.028187556399239436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,10240,0.02427822185887231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,8192,0.015581332974963717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,7168,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,8192,0.02129688858985901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,6144,0.012816000315878125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,7168,0.020139555136362713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,5120,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,6144,0.018927999668651156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,4096,0.009509333305888707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,5120,0.017829333742459614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,3584,0.009036444127559662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,4096,0.01585422290696038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,3072,0.008150222400824228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,3584,0.015544889701737298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,3072,0.014212444424629211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,2560,0.00776533368561003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,2048,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,2560,0.013845333622561561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,1536,0.004979555391603046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,2048,0.013184000220563678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,1024,0.004262222184075249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,1536,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,768,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,1024,0.01239999963177575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,768,0.012154666913880242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,512,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,256,0.003549333247873518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,512,0.01180622229973475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,128,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,256,0.011595555477672152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,64,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,128,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,4096,32,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,64,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,65536,0.08923288848665024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,4096,32,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,51200,0.06839111116197374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,65536,0.11009511020448472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,51200,0.0890408886803521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,16384,0.027105778455734253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,12288,0.02198311189810435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,16384,0.03330844309594896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,10240,0.019561777512232464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,12288,0.027757333384619817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,8192,0.014911111858155994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,10240,0.02360533343421088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,7168,0.01350400017367469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,8192,0.020624000165197585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,6144,0.012164444559150271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,7168,0.019875556230545044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,5120,0.010814222196737925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,6144,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,5120,0.017303110824690927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,4096,0.009289777941173976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,4096,0.015573332707087198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,3584,0.008757332960764566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,3584,0.014879110786649915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,3072,0.008068444828192392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,3072,0.014106666048367819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,2560,0.007270221908887227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,2560,0.014008000493049622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,2048,0.005951111101441913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,2048,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,1536,0.0052720002002186244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,1536,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,1024,0.004584889031118817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,1024,0.012113778127564324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,768,0.00424888895617591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,768,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,512,0.003938666648334927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,512,0.012177777787049612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,256,0.0035048888789282907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,128,0.0032622222271230486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,256,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,64,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,128,0.010818666881985135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3584,32,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,64,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,65536,0.07352799839443631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3584,32,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,51200,0.05931911203596327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,65536,0.10762755738364325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,16384,0.02569777766863505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,51200,0.08645244439442952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,16384,0.03195644418398539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,12288,0.020169778002632987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,10240,0.018435556027624343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,12288,0.026684444811609056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,8192,0.014287999934620328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,10240,0.022987556126382615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,7168,0.013075555364290873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,8192,0.02050844497150845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,6144,0.01183555523554484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,5120,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,6144,0.018038221531444125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,4096,0.008760000268618265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,5120,0.016748444901572335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,7168,0.019291554888089497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,3584,0.008383110993438298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,3072,0.007904888855086433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,4096,0.014674666855070325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,2560,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,3072,0.013872000078360239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,3584,0.014205333259370593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,2048,0.0058986664646201665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,1536,0.004932444542646408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,2560,0.013352889153692456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,1024,0.004310222135649787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,2048,0.012797333300113678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,1536,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,768,0.004233777936961916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,1024,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,512,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,256,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,768,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,128,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,512,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,64,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,256,0.011128000087208219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,128,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,64,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,65536,0.06509777572419909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,3072,32,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,51200,0.05152444375885857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,65536,0.10406488842434353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,16384,0.02159022291501363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,51200,0.08426044384638469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,12288,0.0185208883550432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,12288,0.02644088864326477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,10240,0.01680888897842831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,3072,32,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,8192,0.014465777410401238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,10240,0.022572444544898138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,7168,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,8192,0.02010933392577701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,6144,0.011132444772455426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,7168,0.0192248887485928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,5120,0.010037333601050908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,16384,0.030568000343110826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,6144,0.017826666434605915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,4096,0.008732444710201686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,5120,0.016593777471118502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,3584,0.008391111261314815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,3072,0.0075839997993575195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,3584,0.014575110541449653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,2560,0.0063795554969045855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,4096,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,3072,0.013506666653686099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,2048,0.005573333137565189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,1536,0.004954666727119022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,2560,0.013787555197874704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,1536,0.012498666842778524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,1024,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,1024,0.011976000335481433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,768,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,768,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,512,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,2048,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,256,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,512,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,128,0.0029484445436133277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,256,0.011131555669837527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,64,0.003185777821474605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,128,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,64,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2560,32,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,65536,0.05396977729267544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2560,32,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,51200,0.04344888859324985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,65536,0.10108177529440986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,51200,0.08127999967998929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,16384,0.018481777773963082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,12288,0.014501333236694336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,16384,0.030219554901123047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,10240,0.013711111413107978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,12288,0.02588533361752828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,8192,0.012530666258600024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,10240,0.02215911116864946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,7168,0.011843555503421359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,8192,0.019853333632151287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,7168,0.0184879998366038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,6144,0.010864888628323873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,5120,0.00980888886584176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,6144,0.016917333006858826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,4096,0.008301333420806462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,5120,0.015464888678656684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,3584,0.007679999702506595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,4096,0.014469333820872836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,3072,0.006790222393141852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,3072,0.013516444298956128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,3584,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,2560,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,2048,0.0052728888889153796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,2048,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,1536,0.004574222283230888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,1024,0.004233777936961916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,1536,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,1024,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,768,0.004027555386225383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,512,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,768,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,256,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,512,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,128,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,256,0.011330666641394297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,64,0.003173333282272021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,128,0.011281777587201862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,2048,32,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,64,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,65536,0.0424124432934655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,32,0.011198221809334226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,51200,0.03429155548413595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,65536,0.09861422247356838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,16384,0.015772443678643968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,51200,0.0791520012749566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,16384,0.029827555020650227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,12288,0.01260088880856832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,10240,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,12288,0.02570666703912947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,8192,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,10240,0.022006221943431433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,7168,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,8192,0.018607111440764535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,6144,0.010674667027261523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,7168,0.017874666386180453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,2048,2560,0.013253333667914072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,5120,0.009157333109113906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,4096,0.007648000286685095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,6144,0.016186666157510545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,3584,0.0069191112286514705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,5120,0.015203555425008139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,4096,0.014279110564125909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,3584,0.014118222726715935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,2560,0.0058986664646201665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,3072,0.013537777794731988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,2048,0.004980444494220945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,2560,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,1536,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,2048,0.01276266657643848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,1024,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,1536,0.012707555459605323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,768,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,1024,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,512,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,768,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,256,0.0032720000793536506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,512,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,128,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,3072,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,256,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,128,0.011360888679822287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1536,32,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,64,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,65536,0.03048266636000739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1536,32,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,51200,0.025422222084469263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,65536,0.09575911362965901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,16384,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,51200,0.07630133628845215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,12288,0.01332622187005149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,16384,0.029831111431121826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,10240,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,12288,0.024086222052574158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,8192,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,10240,0.02030933399995168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,7168,0.009745777481132084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,8192,0.018276444739765592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,6144,0.009031111167536842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,7168,0.01740622189309862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,5120,0.008664888640244802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,6144,0.016571554872724745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,4096,0.007616889145639207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,5120,0.01517688896920946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,3584,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,4096,0.014249778456158109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,3072,0.006591111007663939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,3584,0.013966222604115805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,2560,0.005893333504597346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,3072,0.013432000246312885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,2048,0.005251555393139522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,2048,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,1536,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,1536,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,1024,0.0041911109454101985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,1024,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,768,0.0038515557017591265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,768,0.012033777932325998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,512,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,2560,0.013163555827405719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,256,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,512,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,256,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,128,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,64,0.0028320000403457214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,128,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,1024,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,64,0.011134222149848938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,65536,0.02404800057411194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,1024,32,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,51200,0.021337777376174927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,65536,0.09388711055119832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,16384,0.012466666599114736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,51200,0.0748515526453654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,12288,0.01070577816830741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,16384,0.029066665305031672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,10240,0.009896888501114314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,10240,0.020484444167878892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,8192,0.008636444807052612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,8192,0.0184871107339859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,7168,0.008346666892369589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,7168,0.01756088932355245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,6144,0.007616000043021307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,6144,0.016539555456903245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,12288,0.023767999476856653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,5120,0.00796622203456031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,4096,0.007449777589903937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,5120,0.015136000182893542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,3584,0.006660444455014334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,4096,0.014106666048367819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,3072,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,2560,0.005615111026499007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,3072,0.013419555293189155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,2048,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,2560,0.013455111119482251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,1536,0.004584889031118817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,2048,0.012741333080662621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,1024,0.003939555750952827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,1536,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,768,0.0036924444139003754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,1024,0.012076444096035428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,512,0.0032622222271230486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,768,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,512,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,256,0.003597333199448056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,128,0.0031386667655573953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,256,0.011361777782440186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,3584,0.013807999591032663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,64,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,768,32,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,64,0.011015999648306104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,65536,0.018373333745532565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,32,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,51200,0.015391111373901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,65536,0.09238221910264756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,16384,0.010492444038391113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,51200,0.07332622342639498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,12288,0.009027555584907532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,16384,0.027160889572567407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,10240,0.009205333060688442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,12288,0.023132445083724126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,10240,0.02033955521053738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,8192,0.008364444805516137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,8192,0.018231999542978074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,768,128,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,7168,0.007628444168302748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,6144,0.009367999931176504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,7168,0.017628444565667045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,6144,0.01621955633163452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,5120,0.008416888614495596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,4096,0.007369777394665613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,5120,0.01516177753607432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,3584,0.006671111202902264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,4096,0.014559111661381192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,3584,0.01386044422785441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,3072,0.006281777802440856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,2560,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,3072,0.01350400017367469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,2048,0.00527555536892679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,2560,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,2048,0.01257955531279246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,1536,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,1024,0.004087111188305749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,1536,0.012525333298577202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,768,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,1024,0.012177777787049612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,768,0.012058667010731168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,512,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,256,0.0030702222138643265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,512,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,128,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,256,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,64,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,128,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,512,32,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,64,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,512,32,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,65536,0.01479111115137736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,51200,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,65536,0.08930577834447224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,16384,0.009838222629494136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,51200,0.07054489188724093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,16384,0.02715555495686001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,12288,0.014190221826235453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,10240,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,12288,0.022745778163274128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,8192,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,10240,0.02037244372897678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,7168,0.009011555876996782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,8192,0.017937777770890128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,6144,0.008361777497662438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,7168,0.017662222186724346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,6144,0.015992000699043274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,5120,0.00831911133395301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,4096,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,5120,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,3584,0.006706666615274217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,4096,0.014454222387737699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,3072,0.005946666830115848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,3584,0.013551111022631327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,2560,0.005683555371231503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,3072,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,2048,0.004936888813972473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,2560,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,1536,0.004553777890072929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,2048,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,1536,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,1024,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,768,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,1024,0.01184177812602785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,512,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,768,0.012075555821259817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,512,0.011364444262451597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,256,0.00313688897424274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,128,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,256,0.010889777706729041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,64,0.0028159999185138275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,128,0.011347555451922946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,256,32,0.0028080000645584534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,64,0.01071733319097095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,65536,0.014856888188256158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,256,32,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,51200,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,65536,0.08814399772220188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,16384,0.010699555277824402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,51200,0.06914844115575154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,12288,0.009364444348547194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,12288,0.023375110493765935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,10240,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,10240,0.02072622213098738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,8192,0.008085333638721043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,16384,0.026912000444200303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,8192,0.018440888987647164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,7168,0.008626666333940294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,6144,0.00832444429397583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,7168,0.017584888471497428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,6144,0.016057777735922072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,5120,0.007639110916190677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,4096,0.007287110719415877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,5120,0.014914666612943014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,3584,0.006700444552633498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,4096,0.014507555299335055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,3584,0.01423466702302297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,3072,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,3072,0.013802666631009845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,2560,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,2560,0.013536888692114087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,2048,0.004967999955018361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,2048,0.01297688898113039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,1536,0.004576000074545543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,1536,0.01276266657643848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,1024,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,1024,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,768,0.003597333199448056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,768,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,512,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,512,0.012143111063374413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,256,0.0030951110853089225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,128,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,128,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,64,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,64,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,128,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,128,32,0.011809777882364062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,65536,0.01588444411754608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,65536,0.0877608921792772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,51200,0.014569777581426831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,16384,0.009129777550697327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,51200,0.06808711422814263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,12288,0.009418666362762451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,16384,0.027224888404210407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,10240,0.008462222086058723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,12288,0.022993778189023335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,8192,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,10240,0.020452444752057392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,7168,0.007663110891977946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,6144,0.008039111064540016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,8192,0.018407111366589863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,5120,0.007633777956167857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,6144,0.016302223006884258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,4096,0.006664888726340399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,7168,0.017051556044154696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,5120,0.015233778291278414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,3584,0.006342222293217977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,3072,0.007102222078376346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,4096,0.014194667339324951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,2560,0.0063235556913746735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,3584,0.014479110638300577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,2048,0.005933333188295364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,3072,0.013240888714790344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,2560,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,1536,0.004308444344335132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,1024,0.0038373333712418876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,2048,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,768,0.0034977777136696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,1536,0.012465777496496836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,512,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,768,0.012141333685980903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,256,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,512,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,1024,0.011847111086050669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,128,0.0026835555003749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,256,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,64,32,0.002526222210791376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,64,0.011054221954610614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,128,0.010576888918876648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,65536,0.01759288873937395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,51200,0.014544000228246054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,64,32,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,16384,0.009034666750166152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,51200,0.06713777780532837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,65536,0.085825781027476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,12288,0.00942311104800966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,10240,0.008399999803966945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,16384,0.027448000179396734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,8192,0.00795555528667238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,10240,0.02036622166633606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,12288,0.022800000177489385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,7168,0.0074471111098925276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,8192,0.018290667070282828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,7168,0.017130666308932833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,6144,0.01645066671901279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,5120,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,4096,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,5120,0.015240889456537036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,3584,0.006630222416586346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,4096,0.014089778065681458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,3072,0.0062773335311147906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,3584,0.014228444960382251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,2560,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,3072,0.013148444394270578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,6144,0.008031110796663497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,2048,0.005659555395444234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,2560,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,1536,0.004297777596447202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,2048,0.012646222280131446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,1536,0.012744888663291931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,768,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,1024,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,768,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,512,0.0034302222645945023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,512,0.01182933317290412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,256,0.003169777699642711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,1024,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,256,0.01148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,128,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,64,0.0026266665922270883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,128,0.011249778171380361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,64,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,113,32,32,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,113,32,32,0.010768888725174798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,51200,0.6364391114976671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,51200,1.08934752146403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,16384,0.3672791057162815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,16384,0.21219021744198271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,12288,0.268796443939209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,12288,0.1633946630689833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,16384,0.2667377789815267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,10240,0.23165334595574272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,10240,0.138230217827691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,12288,0.2084897756576538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,8192,0.19143554899427626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,8192,0.11052355501386855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,10240,0.17784178256988525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,7168,0.1619582176208496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,51200,0.7805884679158529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,7168,0.09963377979066636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,8192,0.14737244447072348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,6144,0.14303733242882624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,6144,0.08723555670844184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,5120,0.11830843819512261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,7168,0.13298132684495714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,5120,0.07502577702204387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,6144,0.11824711163838704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,4096,0.0975688894589742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,4096,0.06266311142179701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,3584,0.08456266588634914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,5120,0.10292177730136448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,4096,0.08896533648173015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,3584,0.05647644731733534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,3072,0.07347999678717719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,3072,0.05134133497873942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,2560,0.06253511375851102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,3072,0.07471377982033624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,3584,0.08163644207848443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,2048,0.05209155546294319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,2560,0.04501422246297201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,2560,0.06691377692752414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,2048,0.039162665605545044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,1536,0.03945511248376634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,2048,0.061312887403700084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,1024,0.027792887555228338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,1536,0.033165332343843244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,1536,0.052314665582444936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,768,0.022081777453422546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,1024,0.027294221851560805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,1024,0.04564355479346382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,768,0.02269866731431749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,512,0.015650666422314115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,512,0.019736889335844252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,768,0.04226400123702156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,512,0.03644444545110067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,256,0.010462221999963125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,128,0.00775111135509279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,256,0.018225777480337355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,256,0.032255109813478254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,128,0.01684799955950843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,64,0.00775911079512702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,65536,32,0.00832177781396442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,65536,128,0.02803200152185228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,64,0.019301333361201815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,65536,32,0.020995555652512446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,65536,0.6539111137390137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,65536,1.0636186599731445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,51200,0.8384408950805664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,51200,0.5098622110154893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,16384,0.2843608856201172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,16384,0.17165511184268525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,12288,0.21089600192175972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,12288,0.1331253316667345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,16384,0.2335724565717909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,10240,0.16947555541992188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,12288,0.18266044722663033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,10240,0.11225956016116673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,8192,0.1414453321033054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,8192,0.09071111016803318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,10240,0.15629422664642334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,7168,0.11951733960045709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,7168,0.08172888888253106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,8192,0.1301893393198649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,65536,0.8609200053744845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,6144,0.10552089081870185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,51200,0.6753022405836316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,6144,0.07203555769390531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,5120,0.08777778016196357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,7168,0.11717066499922012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,6144,0.10508622063530816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,4096,0.0717537800470988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,5120,0.06207288636101616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,4096,0.05253155695067512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,3584,0.06296444601482816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,5120,0.09177244371838039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,4096,0.07971822553210788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,3072,0.0547342234187656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,3584,0.04764622118737963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,3584,0.07309510972764757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,2560,0.04699200060632494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,3072,0.043455110655890576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,3072,0.06787644492255317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,2560,0.03835466504096985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,2048,0.039270222187042236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,2560,0.06124799781375461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,1536,0.02959288822280036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,2048,0.05565510855780708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,1536,0.02908888790342543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,1024,0.021324444148275588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,1536,0.04948888884650337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,1024,0.02269511090384589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,768,0.016536000702116225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,1024,0.043507556120554604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,768,0.020460445019933913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,512,0.012246222131782107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,768,0.0388995541466607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,512,0.018188445104493033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,512,0.03571288784344991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,256,0.008727110922336578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,2048,0.0339955555068122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,256,0.030777778890397813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,256,0.016520000166363187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,128,0.00562755556570159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,64,0.0052844443255000645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,128,0.01587022178702884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,51200,128,0.028175112273957994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,51200,32,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,64,0.01752000053723653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,51200,32,0.019554666346973844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,65536,0.33333423402574325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,65536,0.26584000057644314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,51200,0.25942222277323407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,51200,0.2096773386001587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,16384,0.09237688779830933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,16384,0.0755902197625902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,16384,0.08702133099238078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,12288,0.06910222106509738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,65536,0.27227732870313853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,51200,0.21413866678873697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,12288,0.06283733579847547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,10240,0.057778669728173145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,10240,0.05266133281919691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,12288,0.0739368862575955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,8192,0.05066133207745022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,10240,0.06276977724499173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,7168,0.040998223755094744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,8192,0.04464177621735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,8192,0.05352266629536947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,7168,0.04052177733845181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,6144,0.03504800134234958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,6144,0.03487288951873779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,7168,0.049654222197002836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,5120,0.030031111505296495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,6144,0.045956446064843066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,4096,0.024596444434589807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,5120,0.03122399912940131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,5120,0.04169777697987027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,4096,0.026504889130592346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,3584,0.02221333318286472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,4096,0.03860977623197768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,3072,0.019539554913838703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,3584,0.024618667032983568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,3584,0.036347554789649114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,2560,0.016546666622161865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,3072,0.034304890367719866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,3072,0.022615111536449854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,2048,0.01401511165830824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,2560,0.02071911096572876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,2560,0.032446222172843084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,1536,0.011069333387745751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,2048,0.030619555049472388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,2048,0.01905866629547543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,1024,0.00903911143541336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,1536,0.017255110873116385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,1536,0.029288887977600098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,768,0.007349333001507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,1024,0.015998222761683993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,1024,0.02723466687732273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,512,0.006544000158707301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,768,0.014505777094099255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,768,0.026403556267420452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,256,0.004219555606444676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,512,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,512,0.023335999912685815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,128,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,256,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,128,0.012518222133318583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,128,0.02033422225051456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,64,0.0036799998746977914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,16384,32,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,64,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,16384,32,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,65536,0.24136622746785483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,65536,0.2138542201783922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,16384,256,0.020630222227838304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,51200,0.18700977166493735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,51200,0.16747821701897514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,65536,0.2347946696811252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,16384,0.06952266560660468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,51200,0.18656622038947213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,16384,0.063137776321835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,12288,0.04974044362703959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,12288,0.05065244436264038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,16384,0.07756711377037896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,10240,0.04303644431961907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,12288,0.06456444660822551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,10240,0.042147556940714516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,8192,0.035197334157096014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,10240,0.05638133154975044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,8192,0.03557777735922072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,7168,0.031129777431488037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,8192,0.04946133494377136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,7168,0.03264000018437704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,6144,0.027642667293548584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,7168,0.04644533329539829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,6144,0.029543111721674602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,5120,0.024099555280473497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,6144,0.04272088739607069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,5120,0.025757332642873127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,4096,0.019506666395399306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,5120,0.03986577855216132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,4096,0.02241955531968011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,4096,0.03629866573545668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,3584,0.017702221870422363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,3584,0.021451556020312842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,3584,0.03426933288574219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,3072,0.015412444869677225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,3072,0.020086222224765353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,2560,0.01346666696998808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,3072,0.03292177783118354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,2560,0.018862222631772358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,2560,0.03152622116936578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,2048,0.01148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,2048,0.029896001021067303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,2048,0.0181022220187717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,1536,0.009680888719028896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,1536,0.016516443755891588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,1536,0.028006222512986925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,1024,0.007914666500356462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,1024,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,1024,0.02574933403068119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,768,0.007231111327807109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,768,0.023749333288934495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,768,0.01351466692156262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,512,0.004896000027656555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,512,0.022295999858114455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,512,0.013023999830087027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,256,0.004223111189073986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,256,0.02069244450993008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,256,0.01238044434123569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,128,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,128,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,12288,128,0.018926221463415358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,64,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,12288,32,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,64,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,12288,32,0.013784888717863293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,65536,0.20300977759891084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,51200,0.1572399934132894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,65536,0.20778222878774008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,51200,0.1645466619067722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,16384,0.058331555790371366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,16384,0.06093955702251858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,51200,0.1791235605875651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,65536,0.2225395573510064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,12288,0.04223555657598707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,12288,0.04896533489227295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,16384,0.0736266639497545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,10240,0.036004443963368736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,12288,0.062237335575951465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,8192,0.029871109459135268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,10240,0.04119999872313605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,10240,0.05464977688259549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,7168,0.026869333452648584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,8192,0.034959110948774554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,8192,0.0472871098253462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,7168,0.031975110371907554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,6144,0.023614222804705303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,7168,0.044488890303505786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,6144,0.02849688794877794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,5120,0.02038755516211192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,6144,0.04134666588571336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,4096,0.01625955601533254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,5120,0.02474311159716712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,5120,0.03857422206136916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,3584,0.014715555641386243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,4096,0.0351164440313975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,4096,0.02221155497762892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,3584,0.02128977742460039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,3584,0.033653332127465144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,3072,0.020101333657900494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,3072,0.0326533317565918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,2560,0.011692444483439127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,2560,0.018738667170206707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,2560,0.031009776724709406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,2048,0.009780444204807281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,3072,0.013208000196350945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,2048,0.017530666457282174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,1536,0.00851466672288047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,2048,0.030226667722066242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,1536,0.015751111838552687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,1024,0.006797333144479328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,1536,0.028509332074059382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,1024,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,768,0.0052800000541739995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,768,0.013410666750537025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,768,0.02380888826317257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,512,0.004597333156400257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,512,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,256,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,1024,0.024440889557202656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,512,0.02336888843112522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,256,0.012308444413873883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,128,0.0037377778854635027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,256,0.021307556165589228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,10240,128,0.019732443822754752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,64,0.00328533330725299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,128,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,10240,32,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,64,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,65536,0.17977689372168648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,10240,32,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,65536,0.16258222526974148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,51200,0.1293368869357639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,65536,0.201493329472012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,16384,0.05000533329115974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,51200,0.13972711563110352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,51200,0.16269777880774602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,12288,0.037802666425704956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,16384,0.0683751106262207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,12288,0.04084533452987671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,10240,0.032961779170566134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,12288,0.05833244323730469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,10240,0.03439999951256646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,8192,0.02799288762940301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,10240,0.05133688780996534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,8192,0.028911110427644517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,7168,0.023971555961502924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,16384,0.050075554185443454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,8192,0.045639998382992215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,7168,0.02681066592534383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,7168,0.04247377647293938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,6144,0.023752000596788194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,5120,0.018253333038753934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,6144,0.03997777899106344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,5120,0.021785777476098802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,4096,0.015000000596046448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,5120,0.03667111198107401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,4096,0.020005333754751418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,3584,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,4096,0.03445511062939962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,3584,0.018608889646000333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,6144,0.021368889345063105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,3584,0.03305866652064853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,3072,0.0321066677570343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,3072,0.018033777674039204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,2560,0.010119111173682744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,2560,0.01718133356836107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,2560,0.029665778080622356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,2048,0.0092284447617001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,2048,0.015499555402331881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,1536,0.007738666401969061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,3072,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,1536,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,1536,0.02758666707409753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,1024,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,1024,0.024400000770886738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,768,0.004972444640265571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,2048,0.029994666576385498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,768,0.012840888566441007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,512,0.004292444636424382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,768,0.023930667175187006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,512,0.012129777835475074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,256,0.003922666526503033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,1024,0.0063768890168931745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,512,0.023756444454193115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,128,0.0038622220357259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,256,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,256,0.021294222937689886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,64,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,8192,128,0.019662222928471036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,128,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,8192,32,0.004217777815130022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,64,0.01183377785815133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,65536,0.16480444537268743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,8192,32,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,51200,0.12611467308468288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,65536,0.15913155343797472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,51200,0.1283528937233819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,16384,0.04261599977811178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,65536,0.1997306611802843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,16384,0.049044443501366504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,51200,0.16077155537075466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,12288,0.03090488910675049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,16384,0.06715466578801473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,12288,0.03947910997602675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,10240,0.026846221751636926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,12288,0.05707910988065931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,10240,0.03276355399025811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,10240,0.05001066790686714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,8192,0.022379555635982092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,8192,0.027506666051016912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,7168,0.019991111424234178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,8192,0.04405511087841458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,7168,0.025450666745503742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,6144,0.01759733259677887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,7168,0.041653331783082746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,6144,0.023565333750512864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,5120,0.015127999915017022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,6144,0.038988444540235735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,5120,0.021599110629823472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,5120,0.03614933292071024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,4096,0.012479999827014076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,4096,0.03422310948371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,4096,0.01961599952644772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,3584,0.01108888867828581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,3584,0.033165332343843244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,3584,0.018558222386572096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,3072,0.010375111467308467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,2560,0.009496000077989366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,3072,0.01753155555990007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,3072,0.03153422143724229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,2048,0.00795377790927887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,2560,0.016721778445773654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,2560,0.030444443225860596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,1536,0.0069324444565508105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,2048,0.01575199928548601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,1536,0.014183999763594734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,1024,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,1536,0.02604533400800493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,1024,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,1024,0.02405599918630388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,768,0.004504000147183736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,768,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,768,0.024381332927280005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,512,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,512,0.012067555553383298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,512,0.022964444425370958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,256,0.003522666792074839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,256,0.021651556094487507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,256,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,128,0.011349333657158745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,128,0.019567999574873183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,64,0.0031822222388452957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,7168,32,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,64,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,7168,32,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,65536,0.13282044728597006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,65536,0.154877331521776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,51200,0.10581155618031819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,7168,2048,0.028228445185555354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,51200,0.12381066216362847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,65536,0.19792266686757407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,16384,0.04284711016549004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,16384,0.04748177859518263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,51200,0.15962044397989908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,16384,0.0658551123407152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,12288,0.03736977775891622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,10240,0.024674667252434626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,12288,0.05645955271191067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,10240,0.0317884451813168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,8192,0.02051733268631829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,10240,0.04996444450484382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,8192,0.026491555902693007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,8192,0.043747554222742714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,7168,0.0181697772608863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,7168,0.0251875552866194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,7168,0.04134133458137512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,12288,0.028112000889248315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,6144,0.016062221593326993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,5120,0.013777777552604675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,6144,0.022845332821210224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,6144,0.03872355487611558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,4096,0.011336888704035016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,5120,0.02095288866096073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,5120,0.03608266512552897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,3584,0.010949333508809408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,4096,0.019300444258583915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,4096,0.0343155562877655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,3072,0.00945688866906696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,3584,0.017969777186711628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,3584,0.03254666593339708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,2560,0.009054222040706212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,3072,0.031457778480317854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,3072,0.01755466726091173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,2048,0.007615110940403408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,2560,0.016167999969588388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,2560,0.029755555921130713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,1536,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,2048,0.014954666296641031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,2048,0.028207110034094915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,1536,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,1024,0.005070222334729301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,1536,0.025376889440748427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,1024,0.02397777802414364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,1024,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,768,0.0041991112132867174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,768,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,768,0.023289778166347083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,512,0.004231111043029361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,512,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,512,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,256,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,256,0.020304888486862183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,128,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,256,0.011511999699804517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,128,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,64,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,6144,128,0.019289778338538278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,64,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,6144,32,0.003277333246337043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,65536,0.11512355009714763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,6144,32,0.011894221935007306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,65536,0.15236443943447536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,51200,0.09162222014533149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,65536,0.19582666291130912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,51200,0.1223120027118259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,16384,0.037696888049443565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,16384,0.046367999580171376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,51200,0.15747555096944174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,12288,0.030466665824254353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,16384,0.06576266553666857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,12288,0.03582133187188043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,10240,0.02626666592227088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,12288,0.055905779202779136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,8192,0.01866755551762051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,10240,0.04970133304595947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,10240,0.03026311265097724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,7168,0.016571554872724745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,8192,0.04288622074657016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,8192,0.02642400066057841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,7168,0.024633778466118708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,7168,0.04092444313897027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,6144,0.014507555299335055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,5120,0.013184888495339287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,6144,0.03845599955982632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,6144,0.02312711046801673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,4096,0.010769777827792697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,5120,0.03556888964441087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,5120,0.0207342223988639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,3584,0.009769777456919352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,4096,0.03360266817940606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,4096,0.01923111081123352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,3072,0.009008000294367472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,3584,0.03225866622394986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,3072,0.03100444542037116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,3072,0.016551110479566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,2560,0.008099555969238281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,2560,0.02796444296836853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,2560,0.015011555618709989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,2048,0.006984000404675801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,2048,0.02676800058947669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,1536,0.005979555762476391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,2048,0.014154665999942355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,1536,0.013493333425786761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,1536,0.02607022225856781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,3584,0.017814222309324477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,1024,0.012959111067983838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,1024,0.023761777414215937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,768,0.004514666481150521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,768,0.012508444488048553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,768,0.0236462222205268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,512,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,1024,0.004976888911591636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,512,0.02342044479317135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,256,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,256,0.02033333314789666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,256,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,128,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,512,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,64,0.003069333318206999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,5120,128,0.01961688862906562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,32,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,64,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,5120,32,0.012737777498033313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,5120,128,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,65536,0.11201511489020453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,51200,0.07578666342629327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,65536,0.19437689251369902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,51200,0.09121333228217231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,65536,0.09707378016577826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,16384,0.030533333619435627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,51200,0.15682578086853027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,16384,0.0655884411599901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,16384,0.03433600068092346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,12288,0.02821777926550971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,12288,0.05551288856400383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,10240,0.022804444034894306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,10240,0.04907466636763679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,10240,0.023919999599456787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,8192,0.015615999698638916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,8192,0.02128355536195967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,12288,0.025860443711280823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,7168,0.014098667436175875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,8192,0.04283377859327528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,7168,0.020054222808943856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,6144,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,7168,0.04073866539531284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,6144,0.018598222070270114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,6144,0.03820533222622342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,5120,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,5120,0.03534222311443753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,5120,0.0177511109246148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,4096,0.009735999835862054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,4096,0.03340177734692892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,4096,0.016199999385409884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,3584,0.009292444421185387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,3584,0.03168977631462945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,3072,0.008442666795518663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,3584,0.015136000182893542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,3072,0.014189332723617554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,3072,0.03080444534619649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,2560,0.00778399987353219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,2560,0.027781334188249376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,2048,0.006548444430033366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,2560,0.014121777481502958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,2048,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,1536,0.005099555684460534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,2048,0.026502221822738647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,1536,0.012774222426944308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,1024,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,1536,0.02514755560292138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,1024,0.012168000141779581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,1024,0.0240346673462126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,768,0.004223111189073986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,512,0.0036480000449551474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,768,0.022876444790098403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,512,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,768,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,256,0.003858666867017746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,512,0.02234399980968899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,256,0.020617778102556866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,256,0.011533333195580376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,128,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,128,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,4096,128,0.019311111834314134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,4096,32,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,65536,0.08265866835912068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,64,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,4096,32,0.011899555722872416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,51200,0.0679031080669827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,65536,0.10986844698588054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,51200,0.08899022473229302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,16384,0.028343111276626587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,65536,0.19339377350277373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,51200,0.15669333934783936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,16384,0.0330079992612203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,12288,0.02216800053914388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,16384,0.06499999761581421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,10240,0.019129777948061626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,12288,0.027479999595218237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,12288,0.05513688921928406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,8192,0.015032000011867948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,10240,0.04882755544450548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,10240,0.023345778385798138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,8192,0.020868443780475195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,7168,0.013471999930010902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,8192,0.042986667818493314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,7168,0.019676445258988273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,7168,0.04039555456903245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,6144,0.018428444862365723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,5120,0.010824888944625854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,6144,0.03809422254562378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,5120,0.01736088924937778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,4096,0.009272889130645329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,5120,0.03533777925703261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,4096,0.015261333849694995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,3584,0.00867377801073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,4096,0.03336711062325372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,3584,0.014170666535695394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,3072,0.008038221961922115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,6144,0.012234666281276278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,3584,0.0306640002462599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,2560,0.007115555306275685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,3072,0.013801777528391944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,3072,0.02949955397182041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,2048,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,2560,0.028206222587161597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,2560,0.013507555756303998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,2048,0.013085333009560903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,1536,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,2048,0.025497777594460383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,1024,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,1536,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,1536,0.02511377798186408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,768,0.004302222281694412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,1024,0.012130666938092975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,1024,0.023735110958417256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,512,0.003914666672547658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,768,0.022672888305452134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,512,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,768,0.012173333101802401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,256,0.0035991109907627106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,512,0.022455111145973206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,256,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,256,0.021408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,128,0.0032720000793536506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,128,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,64,0.0029279999434947968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3584,128,0.019088000059127808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3584,32,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,64,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,65536,0.07240977552202013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3584,32,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,65536,0.1048168871137831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,51200,0.05902489026387533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,65536,0.193375998073154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,51200,0.08616444137361313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,16384,0.025783111651738484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,16384,0.03244355652067397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,51200,0.15624799993303087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,12288,0.02060622142420875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,12288,0.026686223016844854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,12288,0.0551146666208903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,10240,0.018224888377719455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,10240,0.022983999715911016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,16384,0.06448088751898871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,8192,0.015139554937680563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,10240,0.04860799842410617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,8192,0.020643555455737643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,7168,0.014497778481907315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,8192,0.04280622137917412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,7168,0.01941066649225023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,6144,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,7168,0.040399110979504056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,6144,0.01827555563714769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,5120,0.010425777898894416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,6144,0.038049777348836265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,5120,0.01664444473054674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,4096,0.009117333425415887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,5120,0.03567288981543647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,4096,0.014841778410805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,4096,0.03197955422931247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,3584,0.00833155545923445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,3584,0.029903110530641343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,3584,0.014212444424629211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,3072,0.007664889097213745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,3072,0.029061334000693426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,2560,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,3072,0.013768889009952545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,2560,0.013425778183672162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,2560,0.02640977833006117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,2048,0.006148444281684027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,2048,0.02594311038653056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,1536,0.0052062223354975385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,2048,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,1536,0.01278222186697854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,1536,0.0251528885629442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,1024,0.004588444613748127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,1024,0.0237137774626414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,1024,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,768,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,768,0.023725334140989516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,768,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,512,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,512,0.022494221727053326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,512,0.01163822246922387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,256,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,256,0.021309332715140447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,128,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,256,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,128,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,3072,128,0.01889866590499878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,3072,32,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,64,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,65536,0.062353776560889355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,3072,32,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,65536,0.1040159993701511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,51200,0.0503751105732388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,65536,0.1931306653552585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,51200,0.08374133374955918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,16384,0.021643555826610986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,16384,0.03055022160212199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,16384,0.06428888771269056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,12288,0.017895110779338412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,12288,0.025991110338105097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,12288,0.054699556695090405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,10240,0.016493333710564505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,10240,0.02283199959331089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,8192,0.013784888717863293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,51200,0.15608978271484375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,10240,0.048678222629759044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,7168,0.012990222209029727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,8192,0.02027466727627648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,8192,0.042993777328067355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,6144,0.012327999704413943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,7168,0.019271110494931538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,7168,0.04056355688307021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,5120,0.009721777505344814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,6144,0.03847733471128676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,6144,0.017880888448821176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,5120,0.0165964447789722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,4096,0.008752889103359645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,5120,0.03434400094879998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,4096,0.01499288943078783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,3584,0.007888000044557784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,4096,0.031189332405726116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,3584,0.014463111758232117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,3584,0.029930667744742498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,3072,0.0074133334888352295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,3072,0.028640889459186133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,3072,0.013562666873137156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,2560,0.006616000086069107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,2560,0.027071999178992376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,2560,0.013245333400037555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,2048,0.005843555761708154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,2048,0.02682311170630985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,2048,0.01276266657643848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,1536,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,1536,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,1536,0.025776889589097764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,1024,0.004534222185611725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,1024,0.012089778151777057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,768,0.004200889004601372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,1024,0.023687111006842718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,768,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,768,0.022675555613305833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,512,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,256,0.003622222277853224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,512,0.011468444433477191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,512,0.021746666895018682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,128,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,256,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,256,0.020987555384635925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,128,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2560,128,0.020050666398472257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,64,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2560,32,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,65536,0.05102044343948364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,64,0.011227555572986603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2560,32,0.011496889094511667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,51200,0.0414737794134352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,65536,0.10057155291239421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,65536,0.19379999902513292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,16384,0.01828799976242913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,51200,0.07978755235671997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,51200,0.15674400329589844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,16384,0.06398133436838786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,12288,0.014577777849303352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,12288,0.02572711143228743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,10240,0.0136835558546914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,10240,0.02221155497762892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,10240,0.048723555273479886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,16384,0.03020533257060581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,8192,0.011971555650234222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,8192,0.01995377739270528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,8192,0.042675554752349854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,7168,0.011589333415031433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,12288,0.054921779367658824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,6144,0.01089422239197625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,7168,0.04033333394262526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,7168,0.0188746667570538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,5120,0.00999022192425198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,6144,0.017646221650971305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,6144,0.03678933448261685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,4096,0.008450667063395182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,5120,0.03448799914783902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,5120,0.01571999986966451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,4096,0.014260444376203748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,3584,0.007452444069915348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,4096,0.03232355581389533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,3584,0.013824000126785703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,3072,0.007003555695215861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,3584,0.029912001556820337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,2560,0.006047111004590988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,3072,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,3072,0.02862666712866889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,2048,0.0052426668504873914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,2560,0.02681422233581543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,2560,0.01312800000111262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,1536,0.004831111265553368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,2048,0.013073777986897362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,2048,0.025747555825445387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,1024,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,1536,0.025464889076020982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,1024,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,768,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,1536,0.012300444145997366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,1024,0.024393777052561443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,512,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,768,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,768,0.022658665974934895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,512,0.021351110604074266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,512,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,256,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,256,0.02151288919978672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,128,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,2048,128,0.019894222418467205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,64,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,64,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,32,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,2048,32,0.01108888867828581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,2048,256,0.0032337777730491427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,65536,0.04124355647299025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,65536,0.09476444456312393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,65536,0.1935537788603041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,51200,0.032913777563307024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,51200,0.07887111107508342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,16384,0.01516177753607432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,51200,0.15636089113023546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,16384,0.029912001556820337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,12288,0.012881778180599213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,16384,0.06411111354827881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,12288,0.025411556164423626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,10240,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,12288,0.054720001088248364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,10240,0.021756443712446425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,8192,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,10240,0.04853244291411506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,8192,0.018571555614471436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,7168,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,8192,0.04260711206330193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,7168,0.017539555827776592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,6144,0.01053066634469562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,7168,0.03974489039844937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,6144,0.016328000360065036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,5120,0.009126221968068017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,6144,0.03595822387271457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,5120,0.01513155632548862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,4096,0.007672888537247975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,5120,0.03312622176276313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,4096,0.014164444473054675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,3584,0.00701333334048589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,4096,0.030810667408837214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,3584,0.01385600037044949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,3072,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,3584,0.02923111120859782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,3072,0.013563555147912769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,2560,0.005640000104904175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,3072,0.028522666957643297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,2560,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,2048,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,2560,0.026343110534879897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,2048,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,1536,0.004903111192915174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,2048,0.026410667432679072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,1536,0.012455999851226807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,1024,0.004115555435419083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,1536,0.024759110477235582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,768,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,1024,0.02384000023206075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,768,0.01185955521133211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,768,0.023944000403086346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,512,0.0035386664999855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,512,0.011513777905040316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,512,0.022079111801253423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,256,0.003301333429084884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,256,0.021419554948806763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,256,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,1024,0.011765333513418833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,128,0.002969777832428614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,64,0.0029475556479560006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1536,128,0.01833599971400367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1536,32,0.0032666667054096856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,64,0.011131555669837527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,128,0.011013333168294696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,65536,0.03018666638268365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1536,32,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,51200,0.025438222620222304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,65536,0.09517244497934978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,65536,0.19363733132680258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,16384,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,16384,0.029594666428036157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,51200,0.15680444240570068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,12288,0.013241777817408243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,16384,0.0640817748175727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,12288,0.02411288850837284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,12288,0.054506666130489774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,10240,0.011970666547616323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,10240,0.020994666549894545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,8192,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,10240,0.04728000031577217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,51200,0.07599288887447782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,7168,0.00981155534585317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,8192,0.018486221631368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,8192,0.04076977901988559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,6144,0.010018666585286459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,7168,0.038027554750442505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,6144,0.03568622138765123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,5120,0.009019555317031013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,5120,0.015279110934999255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,5120,0.03260533346070184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,4096,0.007742221984598372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,7168,0.017267554998397827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,4096,0.014318222800890604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,3584,0.006983111302057902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,4096,0.030584000878863867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,3584,0.01419911119672987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,6144,0.016217778126398723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,3072,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,3072,0.013491555220550962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,3072,0.028247998820410833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,2560,0.005633777628342311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,2560,0.01311466677321328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,3584,0.029213332467608984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,2048,0.005380444642570283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,2560,0.026146666871176824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,2048,0.012802666260136498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,1536,0.004693333473470476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,2048,0.02643555568324195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,1536,0.012504888905419244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,1024,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,1536,0.024767110745112102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,1024,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,768,0.00398488880859481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,1024,0.023447111248970032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,768,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,512,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,768,0.022749332918061152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,512,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,256,0.0032960000551409195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,512,0.02204977803760105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,256,0.011122666299343109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,128,0.0032755554550223877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,256,0.02033955521053738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,128,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,1024,128,0.0188755558596717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,64,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,1024,32,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,32,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,1024,64,0.010873777998818291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,65536,0.02537244392765893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,65536,0.09300533268186782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,51200,0.02142133315404256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,51200,0.07442577679951985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,51200,0.15622222423553467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,16384,0.012855110896958245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,65536,0.19464266300201416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,16384,0.02887111239963108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,12288,0.011136888629860349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,16384,0.06355200211207072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,12288,0.053202665514416166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,12288,0.02308622168170081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,10240,0.010421333213647207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,10240,0.04600266615549723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,10240,0.020399111840460036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,8192,0.00907555553648207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,8192,0.040228443013297185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,8192,0.01826844447188907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,7168,0.008641777767075432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,7168,0.037816001309288874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,6144,0.008056888977686564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,7168,0.017350221673647564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,6144,0.016182222300105624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,6144,0.03574133250448439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,5120,0.008425777984990014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,5120,0.03264800045225356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,5120,0.015156444576051501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,4096,0.007328000333574083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,4096,0.030604445272021826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,3584,0.007003555695215861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,4096,0.014146667387750415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,3584,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,3584,0.029142220815022785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,3072,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,3072,0.028181334336598713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,2560,0.005648889061477449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,3072,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,2560,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,2048,0.0052177777720822235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,2560,0.02679644525051117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,2048,0.02514044443766276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,2048,0.012655110822783576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,1536,0.004568888909286923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,1536,0.02409066590997908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,1536,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,1024,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,1024,0.022668444448047217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,1024,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,768,0.0038533334930737815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,768,0.022508444057570562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,768,0.011697777443461947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,512,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,512,0.021294222937689886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,256,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,512,0.011500444677140979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,256,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,128,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,128,0.018036444981892902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,64,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,64,0.011013333168294696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,768,32,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,768,32,0.011336888704035016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,768,256,0.020353777541054618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,65536,0.018536888890796237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,65536,0.19423200024498832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,65536,0.08963200118806626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,51200,0.015490666031837463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,51200,0.07184711429807875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,51200,0.15618488523695204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,16384,0.010407110883129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,16384,0.027426666683620874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,16384,0.06355377700593737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,12288,0.009512888888518015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,12288,0.023152000374264185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,12288,0.05264000097910563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,10240,0.009757333331637913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,10240,0.01997244523631202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,8192,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,8192,0.01826133330663045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,10240,0.046078221665488354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,7168,0.009085333181752099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,8192,0.0399982233842214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,7168,0.017456889152526855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,7168,0.03776977790726556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,6144,0.008392000363932716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,6144,0.01627377834584978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,5120,0.008278222547637092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,6144,0.035678221119774714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,5120,0.014834667245546976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,5120,0.03264088763131036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,4096,0.007344889144102733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,4096,0.030233777231640283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,4096,0.01425066590309143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,3584,0.006887110984987683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,3072,0.006228444476922353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,3584,0.029176000091764662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,3072,0.028075555960337322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,3072,0.013733333183659447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,2560,0.005613333235184352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,2560,0.013092444174819522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,2048,0.004990222139490975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,2048,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,2048,0.02674666709370083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,1536,0.004547555413511065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,3584,0.013956444131003486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,1536,0.01238933371173011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,1536,0.02443644404411316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,1024,0.003942222230964237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,1024,0.01200177768866221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,1024,0.024751999311976965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,768,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,768,0.01201777739657296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,768,0.023707555400000677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,512,0.0032808888289663526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,512,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,256,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,256,0.02100533412562476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,2560,0.02676800058947669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,128,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,128,0.017906667457686532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,512,32,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,64,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,65536,0.014818666709793938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,512,32,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,512,512,0.023025777604844835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,51200,0.01350488927629259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,65536,0.1937911113103231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,51200,0.1560924450556437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,51200,0.06975999805662367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,16384,0.009319999979601966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,16384,0.061499555905659996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,12288,0.010071111222108206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,16384,0.0273973329199685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,12288,0.02279288901223077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,10240,0.008336889247099558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,12288,0.05204177896181742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,10240,0.0200284438000785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,8192,0.007648000286685095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,8192,0.017904000149832833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,8192,0.03971822063128153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,7168,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,7168,0.017062221964200337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,7168,0.03772177630000644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,65536,0.09021244446436565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,6144,0.008072000410821702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,10240,0.04609244399600559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,6144,0.016232000456915963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,6144,0.03538489010598924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,5120,0.00796977761718962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,4096,0.007307555940416124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,5120,0.0325564444065094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,4096,0.014430221584108142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,4096,0.03046400017208523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,3584,0.0069057775868309875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,3584,0.013427555561065674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,3584,0.029829333225886028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,3072,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,3072,0.013501333693663279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,2560,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,2560,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,5120,0.01517511076397366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,2560,0.027099554737408955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,2048,0.005135111096832487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,2048,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,2048,0.02683911058637831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,1536,0.004328000048796336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,3072,0.028170665105183918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,1536,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,1024,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,1536,0.024079999989933435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,1024,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,1024,0.0229120006163915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,768,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,768,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,512,0.0033413333197434745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,512,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,512,0.02126666737927331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,256,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,256,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,256,0.01995733380317688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,128,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,128,0.010453333457310995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,768,0.02331200076474084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,64,0.0025537777692079544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,256,128,0.018200889229774475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,256,32,0.002798222212327851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,64,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,65536,0.01478844384352366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,256,32,0.010803555448849997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,65536,0.08782400025261773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,65536,0.19359021716647676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,51200,0.014563555518786112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,16384,0.009313777916961247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,51200,0.06852533419926961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,51200,0.15546845065222845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,16384,0.06135911411709256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,16384,0.027408889598316614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,12288,0.00944622192117903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,10240,0.00870488915178511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,12288,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,12288,0.05200799968507555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,8192,0.007986666427718269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,10240,0.045963555574417114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,10240,0.02035466664367252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,7168,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,8192,0.03976000017589993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,8192,0.018561777141359117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,6144,0.007303111255168915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,7168,0.017306667235162523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,7168,0.03745422098371718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,5120,0.006959999601046245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,6144,0.016521778371598985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,6144,0.03545422355333964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,5120,0.014833778142929077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,4096,0.006974221931563483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,5120,0.033441778686311506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,3584,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,4096,0.013827555709415011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,4096,0.03065511253145006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,3072,0.006704888823959563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,3584,0.02885422110557556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,3584,0.014134221606784396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,2560,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,3072,0.02843377656406826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,3072,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,2048,0.005975111077229182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,2560,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,2560,0.02660088903374142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,1536,0.004261333495378494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,2048,0.012503110700183444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,2048,0.027093332674768236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,1536,0.012156444291273752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,1024,0.003846222327815162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,1536,0.02481066683928172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,1024,0.024454222785101995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,768,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,768,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,768,0.022738666998015508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,512,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,512,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,256,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,256,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,1024,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,128,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,256,0.02035466664367252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,128,0.018570666511853535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,128,0.010450666977299584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,64,0.002562666725781229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,128,32,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,112,128,512,0.023023110296991136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,65536,0.014132445057233175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,64,0.010468444062603844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,128,32,0.010782221953074137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,51200,0.013425778183672162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,16384,0.008442666795518663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,65536,0.08737511105007595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,12288,0.008021333151393467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,51200,0.06822577450010511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,10240,0.007461333440409766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,16384,0.026983999543719824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,8192,0.00702311098575592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,12288,0.023042667243215773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,7168,0.006544000158707301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,10240,0.02013688948419359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,8192,0.01791200041770935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,6144,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,7168,0.01722044414944119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,5120,0.005909333212508096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,6144,0.016010666886965435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,4096,0.006623111251327727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,5120,0.015116444892353482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,3584,0.005920889063013925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,4096,0.013802666631009845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,3072,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,3584,0.013521778086821238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,2560,0.005976888868543837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,3072,0.013150222599506378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,2560,0.012919999659061432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,2048,0.005614222337802251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,1536,0.004366222355100843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,2048,0.01256444470749961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,1024,0.0038622220357259116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,1536,0.01219733307758967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,768,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,1024,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,512,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,768,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,256,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,512,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,128,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,256,0.010769777827792697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,128,0.0107351111041175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,64,32,0.002587555597225825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,64,0.010414222048388587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,65536,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,64,32,0.010709332923094431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,51200,0.013095999757448832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,65536,0.08547466993331909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,16384,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,51200,0.06633066468768649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,12288,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,16384,0.02676800058947669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,12288,0.022981334063741896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,10240,0.007263999846246507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,8192,0.0069617778062820435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,10240,0.020036444067955017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,7168,0.006312888943486744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,7168,0.017254221770498488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,6144,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,6144,0.016149333781666227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,5120,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,5120,0.014939554863505893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,4096,0.006270222365856171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,4096,0.01401600076092614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,3584,0.005982222242487802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,3584,0.013568888935777875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,3072,0.005897777775923411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,2560,0.005920889063013925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,3072,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,2560,0.013090666797426013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,2048,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,1536,0.004243555582231945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,2048,0.01274577776590983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,1024,0.0036959999965296853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,1536,0.012075555821259817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,768,0.003297777846455574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,1024,0.0119982221060329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,8192,0.017934223016103108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,512,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,256,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,768,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,128,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,512,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,256,0.011360000405046674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,64,0.0025884444928831523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,112,32,32,0.002568888788421949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,128,0.010703999963071613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,64,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,112,32,32,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,16384,0.3687022262149387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,51200,0.6293084356519911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,16384,0.20992444621192086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,51200,1.0771715376112196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,12288,0.26919645733303493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,10240,0.2140960031085544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,10240,0.13650222619374594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,8192,0.17553422186109754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,8192,0.10983200205696954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,7168,0.14942399660746256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,7168,0.09880977869033813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,6144,0.13214045100741917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,5120,0.10957422521379258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,6144,0.08656533559163411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,12288,0.1601031091478136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,4096,0.08970755338668823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,5120,0.07464000251558092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,3584,0.07852088742785983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,3072,0.06813066535525851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,4096,0.06190222501754761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,3584,0.05590666664971245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,2560,0.058302223682403564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,3072,0.0509422222773234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,2048,0.04886755678388807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,2560,0.044884443283081055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,2048,0.038888000779681735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,1536,0.03719822234577603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,1024,0.02679288884003957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,1536,0.03293066554599338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,768,0.021347555849287245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,1024,0.03195555673705207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,512,0.014801777071423002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,768,0.027284445034133062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,256,0.009293333523803288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,512,0.020853334003024634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,128,0.008081778056091731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,256,0.017903111047214933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,64,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,128,0.016843555702103507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,65536,32,0.008353778057628209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,64,0.019623999794324238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,65536,32,0.019253333409627277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,65536,0.64738220638699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,51200,0.8359564675225152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,16384,0.28181065453423393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,65536,1.0610853830973308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,12288,0.20957422256469727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,51200,0.5031635496351454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,16384,0.16989155610402426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,12288,0.12946132818857828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,10240,0.16937243938446045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,8192,0.13924445046318903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,7168,0.11881332927280003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,10240,0.1108222206433614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,8192,0.08968889051013523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,6144,0.10502755641937256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,7168,0.08092088831795587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,5120,0.08733955356809829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,6144,0.07155733638339572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,4096,0.07116533650292291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,5120,0.06184977955288357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,3584,0.06257777743869357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,4096,0.05186577637990316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,3072,0.054726223150889076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,2560,0.046478221813837685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,3584,0.04719110992219713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,2048,0.03869244456291199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,3072,0.04305422306060791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,2560,0.03848888807826572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,1536,0.02934933371014065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,2048,0.03360088997417026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,1024,0.021240888370407954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,768,0.016528889536857605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,1024,0.028363555669784546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,512,0.012387555506494312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,1536,0.02915289004643758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,768,0.023494222097926672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,512,0.01904622217019399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,128,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,256,0.01651733285850949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,64,0.004957333207130432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,128,0.015192000402344598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,32,0.00556355549229516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,64,0.015600000818570455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,51200,32,0.016912889149453905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,65536,0.33133422003852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,65536,0.25972623295254177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,51200,256,0.008344000412358178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,51200,0.2591297891404894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,16384,0.09330666727489895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,12288,0.06660533613628812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,51200,0.2038604418436686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,10240,0.056997332308027476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,12288,0.05662844578425089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,16384,0.07441688908471002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,10240,0.049513777097066246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,8192,0.048113776577843555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,7168,0.04071022073427836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,6144,0.03606311149067349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,8192,0.04162933429082235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,7168,0.03843199875619676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,5120,0.030880000856187608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,4096,0.0251857770813836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,6144,0.03406755460633172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,3584,0.022634666826989915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,5120,0.03094044327735901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,4096,0.026399110754330952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,3072,0.01996799972322252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,3584,0.024267555938826665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,2560,0.017253332667880587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,2048,0.013799111048380533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,3072,0.02238666680124071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,2560,0.020629333125220407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,1536,0.011164444188276926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,1024,0.008988444175985124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,2048,0.019119110372331407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,1536,0.017501332693629794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,768,0.007629333270920648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,512,0.005964444329341252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,1024,0.016568889220555622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,768,0.014868444866604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,256,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,128,0.0036755556033717263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,512,0.01423555612564087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,256,0.012536889149083031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,64,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,128,0.012736000120639801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,64,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,16384,32,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,16384,32,0.01219288922018475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,65536,0.24291467666625977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,51200,0.18831200069851348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,65536,0.20770577589670816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,16384,0.06950400273005168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,51200,0.16456977526346842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,16384,0.06149511204825508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,12288,0.04977955420811971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,10240,0.04250577754444546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,12288,0.04719911019007365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,8192,0.034985777404573225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,10240,0.041497776905695595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,8192,0.03509777784347534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,7168,0.03144888745413886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,6144,0.027456889549891155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,7168,0.03230399886767069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,5120,0.02403733299838172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,6144,0.029778665966457788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,4096,0.01943199998802609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,5120,0.026447110705905493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,3584,0.017839110559887357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,4096,0.022460444105996027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,3072,0.015254222684436373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,3584,0.0210524449745814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,2560,0.013420444395807056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,3072,0.020263999700546265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,2048,0.011539555258221097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,2560,0.018876444962289598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,1536,0.009770666559537252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,2048,0.017530666457282174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,1024,0.008079110748238033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,1024,0.015486222174432544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,768,0.007045333584149678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,768,0.014900444282425774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,512,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,256,0.00443733317984475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,512,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,128,0.004004444513056013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,1536,0.01646311084429423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,64,0.0035582222044467926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,128,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,12288,32,0.0035946667194366455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,32,0.012895999683274163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,65536,0.20264355341593424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,256,0.012192000117566852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,65536,0.2015697823630439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,51200,0.1573964489830865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,12288,64,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,16384,0.057987557517157666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,51200,0.1595111158159044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,12288,0.042149331834581166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,16384,0.05951288673612806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,10240,0.03611377875010172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,8192,0.029791110091739233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,12288,0.0456853343380822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,10240,0.04003289010789659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,7168,0.026662222213215295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,6144,0.023423110445340473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,8192,0.034260445170932345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,5120,0.02055022286044227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,7168,0.031347556246651545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,4096,0.016511110795868766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,5120,0.024719999896155462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,6144,0.028899557060665552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,3584,0.0147733340660731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,4096,0.02159999973244137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,3072,0.013241777817408243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,2560,0.011549333731333414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,3072,0.019650666250122916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,2048,0.010080888867378235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,2560,0.01869066721863217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,1536,0.008738666772842407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,2048,0.016899555921554565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,1024,0.007314666277832455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,1536,0.01555288831392924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,768,0.005611555443869696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,1024,0.014277334014574686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,512,0.004641777939266629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,768,0.014246222045686511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,256,0.004017777740955353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,512,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,128,0.0036026665733920205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,256,0.012183999849690331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,64,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,128,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,3584,0.021134222547213238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,64,0.012921777864297232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,65536,0.17928977807362875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,10240,32,0.012535110943847232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,51200,0.13932889037662083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,65536,0.15884444448682997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,16384,0.051464888784620494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,51200,0.12645688321855333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,12288,0.03900622328122457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,10240,32,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,12288,0.03780355718400743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,10240,0.033985777033699885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,10240,0.03328444560368856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,8192,0.028209778997633193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,8192,0.028453333510292902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,16384,0.048977779017554395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,7168,0.02437688906987508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,6144,0.021867555048730638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,7168,0.025666667355431452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,5120,0.018665777312384713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,6144,0.023239110906918842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,4096,0.015503111812803479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,5120,0.021371554997232225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,3584,0.013002666334311167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,4096,0.019343111250135634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,3072,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,3584,0.01863466699918111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,2560,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,2560,0.016913778252071805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,2048,0.00927466650803884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,2048,0.01570933394961887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,1536,0.00797333319981893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,1536,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,1024,0.006334222025341458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,1024,0.013884444203641681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,3072,0.01757066614098019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,768,0.005294222384691238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,512,0.004606222112973531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,768,0.013247999880048962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,512,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,256,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,128,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,128,0.011734222372372946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,64,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,64,0.012034667034943899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,32,0.0031831111345026228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,8192,32,0.012048000262843238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,8192,256,0.004219555606444676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,65536,0.16137066152360705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,51200,0.12475021680196126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,65536,0.15588533878326416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,16384,0.049323555496003896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,51200,0.12274932861328125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,16384,0.04734666811095344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,12288,0.0343191126982371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,10240,0.02904799911710951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,12288,0.03632444474432204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,8192,0.024529778295093115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,10240,0.03222666515244378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,7168,0.02142222225666046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,8192,0.02717599935001797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,7168,0.024792000651359558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,6144,0.01881866653760274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,5120,0.015825778245925903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,6144,0.02296000056796604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,5120,0.0210853334930208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,4096,0.014392889208263822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,3584,0.012525333298577202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,4096,0.018992889258596633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,3072,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,3584,0.018554665976100497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,2560,0.010335110955768162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,3072,0.017273777061038546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,2048,0.008657777474986183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,2560,0.016581333345837064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,1536,0.007275555696752336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,2048,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,1024,0.005912888795137405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,1536,0.014177777700954013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,768,0.004880889008442561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,1024,0.013591110706329346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,512,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,768,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,256,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,512,0.01239377756913503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,128,0.0033688888781600525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,256,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,64,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,128,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,7168,32,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,64,0.012057777908113269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,65536,0.13271378146277532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,7168,32,0.011040888726711273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,51200,0.1051751110288832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,65536,0.1494391096962823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,16384,0.042337778541776866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,51200,0.11873688962724473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,12288,0.03330399923854404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,16384,0.04626577761438158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,10240,0.028226666980319556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,12288,0.03496444556448195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,8192,0.020747555626763236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,10240,0.031122667921913996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,7168,0.018235555953449674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,8192,0.025832000705930922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,6144,0.01608088943693373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,7168,0.02440888848569658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,5120,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,6144,0.022350221872329712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,5120,0.02070577773782942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,4096,0.01202933324707879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,4096,0.019123555885420907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,3584,0.011032000184059143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,3584,0.018245332770877414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,3072,0.010007999837398529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,3072,0.01720266706413693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,2560,0.0090524446633127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,2560,0.015800888339678448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,2048,0.007982222570313348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,2048,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,1536,0.006589333216349284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,1536,0.013769778112570444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,1024,0.005747555444637935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,768,0.004619555754794015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,1024,0.01349244432316886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,512,0.004100444416205089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,512,0.012406222522258759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,256,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,768,0.013719999955760108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,256,0.011889778077602386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,128,0.003469333259595765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,64,0.0034542222403817703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,128,0.01185155577129788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,6144,32,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,64,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,6144,32,0.011807999677128263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,65536,0.11502933502197266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,51200,0.09122311406665379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,65536,0.1468071142832438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,16384,0.03727822171317206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,51200,0.11748711268107097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,12288,0.030905776553683813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,12288,0.0347324444188012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,10240,0.02666933337847392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,10240,0.029694222741656836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,8192,0.018761777215533786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,16384,0.04500177833769056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,7168,0.016513778103722464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,8192,0.02580533259444767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,7168,0.02409155501259698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,6144,0.014704888065656027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,5120,0.01290933373901579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,6144,0.022481777601771887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,4096,0.011569778124491373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,4096,0.018910222583346896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,3584,0.010276444256305695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,3584,0.018058665924602084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,3072,0.009462222456932068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,2560,0.008592888712882996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,3072,0.016551110479566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,2048,0.00740799970097012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,2048,0.014160000615649752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,1536,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,2560,0.014861333701345654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,1024,0.005299555758635203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,1536,0.013853333062595792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,768,0.004615111069546806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,5120,0.02087200019094679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,512,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,1024,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,768,0.013216888739003075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,256,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,128,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,256,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,64,0.0029484445436133277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,128,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,5120,32,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,64,0.012141333685980903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,65536,0.0966977808210585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,32,0.011319110790888468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,65536,0.10698222451739842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,51200,0.08606044451395671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,16384,0.030196444855795965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,5120,512,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,12288,0.02618933386272854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,16384,0.03358844584888882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,10240,0.023044443792766992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,12288,0.026113778352737427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,8192,0.015936000479592215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,10240,0.023365333676338196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,7168,0.01423111061255137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,51200,0.07496622535917494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,8192,0.020710221595234342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,7168,0.019626667102177937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,5120,0.011354666617181567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,6144,0.01868977811601427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,5120,0.01732711162832048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,4096,0.009724444813198512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,3584,0.009185777770148383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,4096,0.016022221909628976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,3072,0.00831288927131229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,3584,0.015277332729763456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,3072,0.013927110367351107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,2560,0.007691555553012424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,2048,0.006603555546866522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,2560,0.013807999591032663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,6144,0.012740444805887012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,1536,0.0052871112194326185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,2048,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,1536,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,768,0.003938666648334927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,1024,0.012503110700183444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,512,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,768,0.012507555385430654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,256,0.0035377778112888336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,512,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,128,0.0032337777730491427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,256,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,64,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,128,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,4096,32,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,64,0.011219555305110084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,65536,0.08185066779454549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,4096,32,0.011671110987663269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,51200,0.0669706662495931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,65536,0.10371377733018662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,16384,0.02786933382352193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,51200,0.08356177806854248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,12288,0.022462222311231825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,16384,0.031664000617133245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,10240,0.019536889261669584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,12288,0.025063999825053748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,8192,0.015791111522250705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,10240,0.022587555978033278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,8192,0.0206666671567493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,7168,0.013399110900031196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,6144,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,7168,0.01926577753490872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,5120,0.010829333629873065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,6144,0.018206222189797294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,4096,0.009358222285906473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,5120,0.017335111896197002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,3584,0.008632000121805403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,4096,0.015104888214005364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,3072,0.009582222335868413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,3584,0.014863111906581454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,2560,0.008669333325492011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,3072,0.013791110780504016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,2048,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,2560,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,1536,0.006493333313200209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,1024,0.005224888937340842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,2048,0.01290933373901579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,1536,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,1024,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,512,0.004221333397759331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,768,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,512,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,256,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,128,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,256,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,64,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,128,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,32,0.0033048888047536216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,64,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,65536,0.07366577784220378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3584,32,0.01149777736928728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3584,768,0.0046862223082118565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,51200,0.05828622314665052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,65536,0.10141866736941868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,16384,0.02581600017017788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,12288,0.02036977807680766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,16384,0.030240890052583482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,10240,0.0181342214345932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,12288,0.024433778391944036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,8192,0.015779554843902588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,10240,0.022343110707071092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,7168,0.014301333162519666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,8192,0.01993333299954732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,6144,0.012053333222866058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,7168,0.01893422173129188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,5120,0.010465777582592435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,6144,0.01791466606987847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,5120,0.016308445069524977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,4096,0.014451555079884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,3584,0.008430221842394935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,3584,0.014200000299347771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,3072,0.007700444095664554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,3072,0.013518222504191928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,2560,0.007333333293596904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,2560,0.013312000367376538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,4096,0.008815999660227034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,2048,0.006260444306664997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,1536,0.005286222116814719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,51200,0.08114666408962674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,2048,0.013174222575293647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,1536,0.012496888637542725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,1024,0.00490133340160052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,768,0.004285333471165763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,512,0.003958222352796131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,1024,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,256,0.00360977773865064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,512,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,128,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,256,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,128,0.010812444819344414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,64,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,3072,32,0.0029813332690132987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,64,0.011139555937714048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,32,0.010479110810491772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,65536,0.06182222233878242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,65536,0.09805955489476521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,3072,768,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,51200,0.050111999114354454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,16384,0.022720888257026672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,51200,0.07866399817996554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,16384,0.02951644526587592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,12288,0.01831733352608151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,10240,0.016286222471131217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,12288,0.02385155525472429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,8192,0.014594667487674289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,10240,0.021963554951879714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,7168,0.01348977784315745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,8192,0.019623999794324238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,6144,0.012487110992272695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,7168,0.0188755558596717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,6144,0.01706133286158244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,5120,0.011485333244005838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,4096,0.008661333057615492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,5120,0.015398222539159985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,3584,0.008079999850855933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,4096,0.014857777290874057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,3584,0.014092443717850579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,3072,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,2560,0.006151999864313338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,3072,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,2048,0.005566222386227713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,2560,0.013345777988433838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,1536,0.004962666581074397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,2048,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,1024,0.00433955548538102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,1536,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,1024,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,768,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,512,0.003966222206751506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,768,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,256,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,512,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,256,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,128,0.010748444332016839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,64,0.0031600000543726813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,64,0.010763555765151978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,32,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2560,32,0.010452444354693094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,65536,0.05074755681885613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2560,128,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,65536,0.09492355585098267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,51200,0.04167200128237406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,16384,0.018127111925019156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,51200,0.07628533575269911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,12288,0.014561777313550314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,16384,0.029296000798543293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,10240,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,12288,0.02342844506104787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,8192,0.012230222423871359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,7168,0.011872889267073737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,8192,0.019197333190176222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,10240,0.021776888105604384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,6144,0.011166222393512726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,5120,0.009845332966910468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,7168,0.017835555805100333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,6144,0.01629866659641266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,5120,0.014833778142929077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,4096,0.008348444269763099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,3584,0.007668444679843054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,3072,0.006767999794748094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,4096,0.01421511173248291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,2560,0.005676444619894028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,3584,0.014089778065681458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,3072,0.013482666677898832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,2048,0.0052328887912962176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,1536,0.004569777597983678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,2560,0.013215111361609565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,2048,0.012806221842765808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,1024,0.004324444466167026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,768,0.003950222084919612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,1536,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,512,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,1024,0.011942221886581846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,256,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,768,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,128,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,512,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,64,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,128,0.011177777416176267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,2048,32,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,64,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,32,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,65536,0.04331555631425646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,51200,0.034605334202448525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,65536,0.09274578094482422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,16384,0.015274667077594332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,51200,0.07415199942058988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,12288,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,16384,0.02905244297451443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,2048,256,0.011035555766688453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,10240,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,8192,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,12288,0.023568888505299885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,10240,0.02118844456142849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,7168,0.010281778044170803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,8192,0.018755555152893066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,7168,0.016959110895792644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,6144,0.010826666322019366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,5120,0.009411555197503831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,6144,0.015847111741701763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,4096,0.0074933336840735535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,5120,0.014956444501876831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,3584,0.0070053330726093715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,4096,0.014128888646761576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,3072,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,3072,0.013400889105266996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,3584,0.013843555417325763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,2560,0.005614222337802251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,2048,0.005236444373925527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,2560,0.013148444394270578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,2048,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,1536,0.004856889032655292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,1024,0.004270222038030624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,1536,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,768,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,1024,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,512,0.0036319999231232535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,768,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,256,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,512,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,256,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,128,0.0031822222388452957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,64,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,128,0.010778666370444827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1536,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,64,0.011159111228254108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,65536,0.030851556195153132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,65536,0.08981421920988296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,51200,0.025773333178626165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,51200,0.07121421893437703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,16384,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1536,32,0.010402667025725046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,16384,0.028541333145565454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,12288,0.013363555901580386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,10240,0.012176888684431711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,10240,0.020271110865804885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,12288,0.023079999619060095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,8192,0.010977778169843884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,7168,0.009757333331637913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,8192,0.017980444762441847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,7168,0.016938666502634685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,6144,0.01051644484202067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,5120,0.008873778084913889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,6144,0.01556622154182858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,4096,0.007663110891977946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,5120,0.014883556299739413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,3584,0.006643555644485686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,4096,0.014118222726715935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,3072,0.006435555716355641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,3584,0.01386844449573093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,3072,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,2560,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,2048,0.006069333189063602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,2560,0.013181333740552267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,1536,0.00526311124364535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,2048,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,1024,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,1536,0.012313777373896705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,768,0.004233777936961916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,1024,0.0118231111102634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,768,0.012072000238630505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,512,0.00388355553150177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,256,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,512,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,128,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,256,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,64,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,128,0.010802666346232096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,1024,32,0.0034622223012977173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,64,0.011136888629860349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,1024,32,0.010072888599501716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,65536,0.024352888266245525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,51200,0.020899555749363367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,65536,0.08790400293138291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,16384,0.012788444757461548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,51200,0.06919911172654894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,12288,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,16384,0.027655111418830022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,10240,0.01012711144155926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,12288,0.021716444028748408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,10240,0.01961599952644772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,8192,0.008730666504965888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,7168,0.008685333033402761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,8192,0.01789422167672051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,6144,0.008020444048775567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,7168,0.016885333591037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,5120,0.008049777812427944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,6144,0.015763555963834126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,4096,0.007332444190979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,5120,0.014898666077189975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,4096,0.014000000225173103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,3584,0.0069857777820693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,3584,0.013791999883121915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,3072,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,3072,0.013456000222100152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,2560,0.005990222096443176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,2560,0.013190222283204397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,2048,0.0052737775776121355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,2048,0.012511110968059964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,1536,0.004634666774008009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,1536,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,1024,0.003956444561481476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,1024,0.01181244436237547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,768,0.0036071112586392295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,768,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,512,0.0032595555401510666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,512,0.011510222322411008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,256,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,256,0.010793777803579966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,128,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,128,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,64,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,64,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,768,32,0.0028115556471877625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,768,32,0.010148444109492833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,65536,0.01811555524667104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,65536,0.08617689212163289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,51200,0.015487111277050443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,51200,0.0665022201008267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,16384,0.010453333457310995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,16384,0.026155556241671245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,12288,0.00943911075592041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,12288,0.021494223011864558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,10240,0.009367999931176504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,10240,0.019588443968031142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,8192,0.008362666600280339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,8192,0.017834666702482436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,7168,0.009312888814343346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,7168,0.016911110944218107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,6144,0.009696889254781935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,6144,0.015543111496501498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,5120,0.008443555898136562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,4096,0.007337777978844113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,5120,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,3584,0.0069653333889113525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,4096,0.01426577733622657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,3072,0.006076444354322221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,3584,0.013891556196742587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,2560,0.006152888966931238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,3072,0.01349688900841607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,2560,0.01310400002532535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,2048,0.004950222041871813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,1536,0.004612444589535396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,2048,0.012535110943847232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,1024,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,1536,0.012477333347002665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,768,0.00361333332127995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,1024,0.011728888584507836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,512,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,768,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,256,0.0031840000301599503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,512,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,128,0.002938666691382726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,256,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,64,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,128,0.010716444088353051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,512,32,0.0027875554644399216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,64,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,512,32,0.01038577738735411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,65536,0.014473777678277759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,51200,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,65536,0.08206578095753987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,16384,0.010082667072614035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,51200,0.06569600105285645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,12288,0.009672889278994666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,16384,0.025783111651738484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,10240,0.008364444805516137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,12288,0.021627555290857952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,8192,0.008015111088752747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,10240,0.019331556227472093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,7168,0.009022222624884712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,8192,0.017980444762441847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,6144,0.00814488861295912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,7168,0.016501333978441026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,5120,0.008372444245550368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,6144,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,4096,0.007317333585686154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,5120,0.014494222071435718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,3584,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,4096,0.014121777481502958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,3072,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,3584,0.013784888717863293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,2560,0.00553955551650789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,3072,0.013365333278973898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,2048,0.004875555634498596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,2560,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,1536,0.004543999830881755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,1024,0.003942222230964237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,2048,0.01255022237698237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,768,0.003510222252872255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,1536,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,512,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,1024,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,768,0.012068444656001197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,256,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,512,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,128,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,256,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,64,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,128,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,256,32,0.002543111021320025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,32,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,65536,0.014779556128713818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,65536,0.08118400308820936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,51200,0.014256000518798828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,51200,0.06471288866466947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,16384,0.009327111144860586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,256,64,0.011028444601429833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,12288,0.009359111388524374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,16384,0.026426666312747534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,10240,0.008707555631796518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,12288,0.021645334031846788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,8192,0.007963555554548899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,10240,0.01997511088848114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,7168,0.008715555899673039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,8192,0.01792533364560869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,6144,0.00797155582242542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,7168,0.017090666625234816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,5120,0.007592000067234039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,6144,0.015788444214397006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,4096,0.0069297779765393995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,5120,0.015169777803950839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,3584,0.00675644435816341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,4096,0.014102222190962898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,3072,0.005964444329341252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,3584,0.014186667071448432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,3072,0.013221333424250284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,2560,0.01309866706530253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,2048,0.004876444323195351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,2048,0.012455999851226807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,1536,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,1536,0.01239111108912362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,1024,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,1024,0.012039999994966718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,768,0.0035662220584021676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,2560,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,512,0.0034462221794658234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,768,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,256,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,512,0.011960888902346293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,128,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,64,0.002526222210791376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,256,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,128,32,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,128,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,65536,0.015509333875444202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,64,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,128,32,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,51200,0.014708444476127625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,16384,0.00904266701804267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,65536,0.08066488636864556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,12288,0.009393778112199571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,51200,0.06333155764473809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,10240,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,16384,0.02576622201336755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,12288,0.02169688873820835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,8192,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,10240,0.01961599952644772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,7168,0.007282666862010956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,6144,0.0070497774415545994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,8192,0.017952000101407368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,7168,0.016885333591037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,5120,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,4096,0.006815111057625876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,6144,0.015652444627549913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,5120,0.015068444940778943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,3584,0.006636444479227066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,3072,0.006346666564544042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,4096,0.01445777714252472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,2560,0.006362666686375936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,3584,0.01387555566098955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,2048,0.006001777946949005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,3072,0.013208889298968844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,2560,0.013429332938459186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,1536,0.004295111116435793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,2048,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,1024,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,768,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,1536,0.012789333032237159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,1024,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,512,0.0032844444115956626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,256,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,768,0.011790222591824003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,128,0.0026195556339290407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,512,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,64,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,256,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,64,32,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,128,0.010766222245163388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,64,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,65536,0.015967110792795818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,64,32,0.011407111254003314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,51200,0.014259555273585849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,65536,0.07939466502931383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,16384,0.00906844437122345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,51200,0.060565332571665444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,12288,0.009073778159088558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,16384,0.02612000041537815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,10240,0.0086666668454806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,12288,0.021332444416152105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,10240,0.0196115556690428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,8192,0.007643555601437886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,7168,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,8192,0.017907554904619854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,6144,0.007033777733643849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,7168,0.01663644446267022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,5120,0.006659555352396435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,6144,0.015882665912310284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,4096,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,5120,0.014862222803963555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,4096,0.014126222994592456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,3584,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,3072,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,3584,0.013815110756291283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,2560,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,3072,0.013200889031092325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,2048,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,2560,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,1536,0.004250666747490565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,2048,0.012372444073359171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,1024,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,1536,0.012413333687517377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,768,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,1024,0.011601777540312873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,768,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,512,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,256,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,512,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,128,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,128,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,64,0.002611555573013094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,256,0.011206222077210745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,97,32,32,0.0025475554996066624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,64,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,97,32,32,0.010476444330480365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,51200,0.6288320223490397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,51200,1.0868026945326064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,16384,0.3674026595221625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,16384,0.2100506623586019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,16384,0.2646088865068224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,12288,0.2696613205803765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,12288,0.16036800543467203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,10240,0.21275644832187227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,10240,0.13624710506863064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,12288,0.20533511373731825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,8192,0.1752986643049452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,8192,0.11013244258032905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,51200,0.7701093355814616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,10240,0.17553777164883086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,7168,0.14743110868665907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,7168,0.09860355324215359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,6144,0.131841778755188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,8192,0.1456257767147488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,6144,0.0865066647529602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,5120,0.10935378074645996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,7168,0.13183644082811144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,5120,0.07449422279993693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,4096,0.08950310945510864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,6144,0.11641777886284722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,4096,0.061920000447167285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,5120,0.10278577937020196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,3584,0.0790497793091668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,3584,0.05600000090069241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,3072,0.06810222069422404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,3584,0.08132977618111505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,3072,0.05073777834574381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,2560,0.05799555778503418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,3072,0.07365244626998901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,2560,0.04455111092991299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,2048,0.04852799905671013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,2560,0.06725511285993788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,4096,0.08846488926145767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,1536,0.03717511230044895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,2048,0.03901688920127021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,2048,0.05902044640647041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,1024,0.026496888862715826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,1536,0.05267288949754503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,1536,0.03289599882231818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,1024,0.02790400054719713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,768,0.020974222156736586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,1024,0.04609333475430807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,768,0.041131556034088135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,512,0.014547554983033074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,768,0.023708444502618577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,256,0.009275555610656738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,512,0.020829333199395072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,512,0.03629511263635423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,256,0.01791822248035007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,128,0.00792266676823298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,256,0.03260533346070184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,64,0.006616888774765863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,128,0.016600888636377122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,64,0.01724088854259915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,65536,32,0.00665244460105896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,65536,128,0.02720355490843455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,65536,32,0.020959110723601446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,65536,0.6460328631930882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,65536,1.057501369052463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,51200,0.8326871130201551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,51200,0.5024435785081651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,65536,0.8494124412536621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,16384,0.2861795425415039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,51200,0.6665146615770128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,16384,0.16954667038387725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,12288,0.2090746694140964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,12288,0.12991644276512995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,16384,0.23099909888373482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,10240,0.16887822416093615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,12288,0.17920622560713026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,10240,0.1105946699778239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,8192,0.1392586628595988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,10240,0.15436444017622206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,8192,0.08966044584910075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,7168,0.11809955702887641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,7168,0.08097333378261991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,8192,0.12820888890160456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,6144,0.10477422343360053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,7168,0.11618755923377143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,6144,0.07125244537989299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,5120,0.08707733286751641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,5120,0.06176444556978014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,6144,0.10368977652655707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,4096,0.07041244374381171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,5120,0.0914577775531345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,4096,0.05213333169619242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,3584,0.0623262193467882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,4096,0.07942933506435819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,3584,0.04725955592261421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,3072,0.05416622095637851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,3584,0.07318222522735596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,3072,0.043046222792731396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,2560,0.046487109528647534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,3072,0.06713066498438518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,2560,0.03815022110939026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,2048,0.03849511014090644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,2048,0.03385600116517808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,2048,0.05426488982306587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,1536,0.029358221424950495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,1536,0.028699556986490887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,1024,0.020988444487253826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,1536,0.04797244403097364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,1024,0.024467556013001338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,1024,0.04211911227968004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,768,0.015917332635985482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,512,0.011991110940774282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,768,0.03827288746833801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,768,0.020803555846214294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,512,0.01864355636967553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,256,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,2560,0.06189155578613281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,512,0.0340302222304874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,128,0.0053431110249625305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,256,0.030047112041049536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,64,0.005251555393139522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,128,0.015478221906556023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,51200,128,0.02609688871436649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,51200,32,0.0052266667286554975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,64,0.017755554782019723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,32,0.01906044450071123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,65536,0.3324497805701362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,65536,0.25911378860473633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,51200,0.2581848833296034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,51200,256,0.016226665841208566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,51200,0.20350578096177843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,16384,0.09100266959932114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,16384,0.07395822472042508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,12288,0.06686666938993666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,16384,0.08398489157358806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,12288,0.05641600158479479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,10240,0.057143999470604785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,10240,0.04937600096066793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,12288,0.06691377692752414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,65536,0.25611289342244464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,8192,0.04738577869203356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,8192,0.04159999887148539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,10240,0.060179558065202504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,8192,0.05260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,7168,0.04064711266093784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,6144,0.03612711032231649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,7168,0.03811733259095086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,7168,0.04819022284613716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,6144,0.0344817770851983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,5120,0.03049066662788391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,6144,0.04495199852519565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,4096,0.024812445044517517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,5120,0.03101955519782172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,5120,0.042019555966059365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,3584,0.02236266599761115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,4096,0.026595556073718604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,4096,0.03850400116708543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,3072,0.019783111082182992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,3584,0.024283554818895128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,3584,0.03577422102292379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,3072,0.022355554832352534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,2560,0.016985777351591323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,51200,0.20180088943905303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,3072,0.03408622079425388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,2048,0.014049778381983439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,2560,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,2560,0.031926221317715116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,2048,0.018889778190188937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,1536,0.011360000405046674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,2048,0.03160533308982849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,1024,0.008613333106040955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,1536,0.017342221405771043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,1536,0.02920711040496826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,768,0.007675555845101674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,1024,0.02730400032467312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,1024,0.016210666961140104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,512,0.006324444380071428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,768,0.023757333556811016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,768,0.014553777045673795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,256,0.0038720000949170855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,512,0.013822221921549903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,512,0.022637334134843614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,128,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,256,0.02271466619438595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,256,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,128,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,64,0.0035822221802340615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,16384,32,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,16384,128,0.019955555597941082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,64,0.013131555583741931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,16384,32,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,65536,0.2678604390886095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,65536,0.207750227716234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,51200,0.20996177196502686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,51200,0.16391555468241373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,16384,0.0748151143391927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,16384,0.061477336618635386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,16384,0.07598221964306302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,65536,0.2240755558013916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,12288,0.05490666627883911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,12288,0.047228445609410606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,12288,0.06090311209360758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,10240,0.046815110577477344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,10240,0.041249778535630964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,8192,0.03920177618662516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,10240,0.054923554261525474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,8192,0.034959110948774554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,51200,0.17802577548556855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,7168,0.03360533383157518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,8192,0.04801422357559204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,6144,0.030103110604816016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,7168,0.04486133323775398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,7168,0.0326222214433882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,5120,0.026280888252788123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,6144,0.04156977931658427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,5120,0.026872000760502283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,6144,0.029425776667065088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,4096,0.021447999609841242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,5120,0.03879377908176846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,3584,0.018929777873886954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,4096,0.03626311156484816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,4096,0.022408000297016565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,3072,0.016813332835833233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,3584,0.02128177715672387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,3584,0.03444000085194906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,2560,0.014764444695578681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,3072,0.03267466690805223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,3072,0.020066666934225295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,2560,0.0317466656366984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,2560,0.019108444452285767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,2048,0.01719199948840671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,2048,0.030378666188981798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,1536,0.010696888797812991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,1536,0.016454221473799813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,1536,0.02776622275511424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,1024,0.008558221989207799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,1024,0.014486221803559197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,1024,0.02681244413057963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,768,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,768,0.02307022280163235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,768,0.013765333427323235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,512,0.004935111022657818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,2048,0.012618666721714867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,256,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,512,0.013171555267439948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,256,0.022635555929607813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,128,0.00361333332127995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,256,0.012230222423871359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,128,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,64,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,128,0.019072888625992667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,12288,32,0.0036311112344264984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,64,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,12288,32,0.01255733354224099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,12288,512,0.022396443618668452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,65536,0.19945422808329263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,51200,0.15938933690388998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,51200,0.15738488568200007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,65536,0.21165066295199922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,65536,0.2134693331188626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,16384,0.06070933077070448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,16384,0.05930311150021023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,51200,0.16998044649759927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,16384,0.07247466511196561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,12288,0.04598222176233927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,12288,0.04550666610399882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,10240,0.03885511226124234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,12288,0.05810578001870049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,10240,0.0398888885974884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,8192,0.0325546662012736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,10240,0.053073777092827686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,8192,0.03397422366672092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,7168,0.028984887732399836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,8192,0.04606399933497111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,7168,0.03109333250257704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,6144,0.025605332520272996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,7168,0.04353244437111748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,6144,0.027627555860413447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,5120,0.02104088829623328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,5120,0.02476177778508928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,6144,0.040165334939956665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,4096,0.018166222506099276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,5120,0.03783199853367276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,4096,0.034684442811542086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,3584,0.015068444940778943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,4096,0.022133333815468684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,3584,0.020928000410397846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,3072,0.01348000019788742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,3584,0.03379199902216593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,3072,0.01962933275434706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,2560,0.01183555523554484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,3072,0.031147556172476873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,2560,0.018464888135592144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,2048,0.010278222461541494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,2048,0.01717333330048455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,2560,0.030943999687830608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,1536,0.009033777647548253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,2048,0.02882666720284356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,1024,0.00720000018676122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,1536,0.027278222971492346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,1536,0.015671999918089975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,1024,0.01406222250726488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,768,0.006121777825885349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,768,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,512,0.0047475554876857335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,512,0.012788444757461548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,512,0.023062222533755835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,256,0.003938666648334927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,1024,0.024101333485709295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,256,0.011873777541849347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,128,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,256,0.020617778102556866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,128,0.011857777833938599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,128,0.01790577835506863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,64,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,10240,32,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,64,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,10240,32,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,65536,0.18227199713389078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,65536,0.16006488270229763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,51200,0.13836977216932508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,10240,768,0.023345778385798138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,51200,0.1266231139500936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,16384,0.0505617790751987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,65536,0.1931084394454956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,16384,0.04863466819127401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,12288,0.036907556984159685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,16384,0.06632977724075317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,12288,0.0376835564772288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,10240,0.031939556201299034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,51200,0.15459911028544107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,10240,0.03338844577471415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,12288,0.05502666698561775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,8192,0.02702399922741784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,8192,0.02823822365866767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,10240,0.04973511232270134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,8192,0.04387288954522875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,7168,0.023303111394246418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,6144,0.020972443951500788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,7168,0.04180711176660326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,6144,0.02293066680431366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,5120,0.017853332890404593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,7168,0.025740444660186768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,6144,0.03857510950830247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,4096,0.014793778459231058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,5120,0.021083555287784998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,4096,0.019630221856964957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,4096,0.033599111768934466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,5120,0.03642755415704515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,3584,0.013247999880048962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,3072,0.01200177768866221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,3584,0.01869777838389079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,3072,0.017834666702482436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,3584,0.03272711237271627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,2560,0.010680889089902243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,3072,0.03150399857097202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,2048,0.009683555199040307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,2560,0.03127200073666043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,2048,0.01576622161600325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,2560,0.016751110553741455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,2048,0.02865511178970337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,1536,0.008194666769769456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,1536,0.014262222581439547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,1024,0.005624889085690181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,1536,0.02680977847841051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,1024,0.013427555561065674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,768,0.004635555462704765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,1024,0.02409422232045068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,768,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,512,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,768,0.023373333944214716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,512,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,256,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,512,0.02242044442229801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,256,0.011830222275522021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,128,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,256,0.020759999752044678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,128,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,64,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,8192,128,0.018386666973431904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,64,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,8192,32,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,8192,32,0.012138666378127204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,65536,0.16448177231682673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,65536,0.15450755755106607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,51200,0.1252666711807251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,51200,0.12230043941073948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,16384,0.046154667933781944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,65536,0.19068888823191324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,16384,0.047352890173594155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,12288,0.03481600019666884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,51200,0.15308088726467556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,16384,0.06522222359975179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,12288,0.036496887604395546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,10240,0.0292871097723643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,12288,0.054083555936813354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,8192,0.024653333756658766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,10240,0.03231466809908549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,10240,0.0495395561059316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,8192,0.02769333289729224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,7168,0.020759999752044678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,8192,0.04346133271853129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,6144,0.018967999352349173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,7168,0.025027554896142747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,7168,0.04087911049524943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,6144,0.024010666542583044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,6144,0.03835733400450812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,5120,0.016523554921150208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,4096,0.014415999253590902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,5120,0.035997334453794695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,4096,0.03381155596839057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,4096,0.0192266669538286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,3584,0.01238844460911221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,3584,0.03287377622392442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,3072,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,3584,0.018254222141371835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,3072,0.017910222212473553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,2560,0.009383999639087254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,3072,0.03058755397796631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,2560,0.016604445046848722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,5120,0.021451556020312842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,2048,0.00833511104186376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,2048,0.02849510974354214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,2048,0.015188443991872998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,1536,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,1536,0.013829333086808523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,1536,0.025256888733969793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,1024,0.005657777604129579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,1024,0.013208000196350945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,1024,0.02444533341460758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,768,0.013047999805874295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,2560,0.029926220575968426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,512,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,768,0.024068444967269897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,512,0.0233324451578988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,256,0.0042168887125121225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,512,0.012155555188655853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,256,0.01204622205760744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,128,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,768,0.004942222187916438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,256,0.02102488941616482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,64,0.0032799999333090256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,128,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,7168,128,0.019274666905403137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,7168,32,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,32,0.012366222010718452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,65536,0.13233689467112222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,7168,64,0.011593778100278644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,51200,0.10276444753011067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,65536,0.1498391098446316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,51200,0.11675200197431777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,16384,0.04150844282574124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,65536,0.18903022342258027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,51200,0.1514542235268487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,12288,0.032187554571363665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,16384,0.06497333447138469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,16384,0.04622488882806566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,10240,0.028226666980319556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,12288,0.03511555658446418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,12288,0.05380710959434509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,10240,0.030896888838873968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,10240,0.0487937761677636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,8192,0.021339555581410725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,8192,0.026556443836953905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,7168,0.018608000543382432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,8192,0.042952001094818115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,7168,0.02461066676510705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,6144,0.01646222174167633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,7168,0.04038577940728929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,6144,0.022553778356975977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,6144,0.03817244370778402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,5120,0.020975111259354487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,5120,0.03577155537075467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,4096,0.012513777448071374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,5120,0.014767999450365702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,4096,0.01904000010755327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,3584,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,4096,0.03378488951259189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,3584,0.01809777816136678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,3584,0.03254222207599216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,3072,0.009970666633711921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,3072,0.01724088854259915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,3072,0.030266665750079687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,2560,0.009399111072222391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,2560,0.030284444491068523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,2560,0.01644088824590047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,2048,0.00813600007030699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,2048,0.028991109795040552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,2048,0.01475288967291514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,1536,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,1536,0.02682222260369195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,1536,0.013981334037250943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,1024,0.0053671110007498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,1024,0.024067555864651997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,768,0.00489688871635331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,768,0.012887111140622033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,768,0.023727999793158636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,512,0.00407644444041782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,512,0.012465777496496836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,512,0.022239110536045496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,256,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,256,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,256,0.022336888644430373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,128,0.0034373334298531213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,1024,0.012975999878512489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,64,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,6144,128,0.01756266587310367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,128,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,6144,32,0.0031031111462248694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,65536,0.11486577987670898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,64,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,65536,0.14405332671271429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,51200,0.09015111128489177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,51200,0.11464444796244304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,65536,0.18676800198025176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,16384,0.037208000818888344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,6144,32,0.012148444851239523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,16384,0.04493777619467842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,51200,0.15020533402760824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,12288,0.029449777470694646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,16384,0.06499555375840929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,10240,0.025960000024901494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,12288,0.03372622198528714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,12288,0.052632000711229115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,8192,0.01898755629857381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,10240,0.04854489035076565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,10240,0.029804444975323145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,7168,0.01628800067636702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,8192,0.04225422276390923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,8192,0.0258933338854048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,7168,0.024068444967269897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,6144,0.014639111028777229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,7168,0.04013777772585551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,5120,0.01294222225745519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,6144,0.022054221895005967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,6144,0.03789155681927999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,4096,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,5120,0.02049422264099121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,5120,0.03576888971858554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,3584,0.010690666735172272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,4096,0.03293955657217238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,4096,0.018931556079122756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,3072,0.009826666778988307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,3584,0.017889777819315594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,3584,0.03187733226352268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,2560,0.008692444198661381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,3072,0.030631999174753826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,3072,0.016932444439993966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,2048,0.007738666401969061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,2560,0.028492444091373022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,2048,0.027811557054519653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,2048,0.014000889327791003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,1536,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,1536,0.02509422269132402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,1536,0.013208000196350945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,1024,0.005281777845488654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,1024,0.023722666833135817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,768,0.004583111239804162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,2560,0.015466666883892484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,768,0.012723555167516073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,768,0.023763555619451735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,512,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,512,0.012149333126015134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,512,0.023056000471115112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,1024,0.012839111189047495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,256,0.0033155555526415506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,256,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,128,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,256,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,128,0.011698666546079846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,64,0.003020444470975134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,5120,128,0.018245332770877414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,5120,32,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,64,0.012156444291273752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,65536,0.09810577498541938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,5120,32,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,65536,0.10579999950197007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,51200,0.07445600297715929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,65536,0.18503467241923013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,51200,0.08553955290052627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,16384,0.029537777105967205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,16384,0.03365688853793674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,51200,0.14928800529903835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,12288,0.02583733366595374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,16384,0.0644213358561198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,12288,0.025750221477614507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,12288,0.052489777406056724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,10240,0.02073777715365092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,10240,0.023306666149033442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,10240,0.047712001535627574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,8192,0.015634665886561077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,7168,0.013977777626779346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,8192,0.04286311070124308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,8192,0.020598222812016804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,6144,0.012865777644846173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,7168,0.019831111033757527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,7168,0.040876444843080305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,5120,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,6144,0.01854044364558326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,6144,0.037408000893063016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,4096,0.009581333233250512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,5120,0.017603556315104168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,5120,0.035375111632876925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,4096,0.016352888610627916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,4096,0.033546666304270424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,3584,0.014910222755538093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,3584,0.031897776656680636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,3072,0.008368888662921058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,3072,0.02878755662176344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,3072,0.015141333142916361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,2560,0.007599111232492659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,2560,0.014168888330459595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,2560,0.028110222684012517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,2048,0.006326222171386083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,2048,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,2048,0.02679466704527537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,1536,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,3584,0.009320889082219865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,1024,0.004587555511130227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,1536,0.024448888169394598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,1536,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,768,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,1024,0.024409777588314478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,1024,0.012502222425407834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,768,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,512,0.003644444462325838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,768,0.023100444012218054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,256,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,512,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,512,0.021712889273961384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,256,0.011152000062995486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,128,0.003272888975010978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,256,0.02198222279548645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,64,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,128,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,4096,32,0.003018666679660479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,4096,128,0.01830933325820499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,64,0.011701333026091257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,65536,0.08136266469955444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,4096,32,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,51200,0.06636266575919257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,65536,0.10309955808851455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,51200,0.08308177524142794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,16384,0.027443556321991816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,65536,0.18525421619415283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,51200,0.14987466070387098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,16384,0.031889776388804116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,12288,0.022031999296612207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,16384,0.06339466571807861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,12288,0.025124443901909724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,10240,0.019907555646366544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,12288,0.05271288752555847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,8192,0.016255110502243042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,10240,0.023009777069091797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,10240,0.047622223695119224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,7168,0.01606044504377577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,8192,0.020597333709398907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,8192,0.04195555713441637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,6144,0.012072000238630505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,7168,0.04048444496260749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,7168,0.01943733294804891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,6144,0.01847644481394026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,5120,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,6144,0.03731288843684726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,5120,0.01722044414944119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,4096,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,5120,0.03556266758177016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,4096,0.015478221906556023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,3584,0.008824889030721452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,4096,0.03269066744380527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,3584,0.014271999398867289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,3584,0.030379556947284277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,3072,0.010108444425794814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,3072,0.029073778125974867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,3072,0.014575999644067554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,2560,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,2560,0.028187556399239436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,2560,0.013249778085284762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,2048,0.0063279999627007385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,2048,0.026740445031060114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,2048,0.012879999975363413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,1536,0.012479999827014076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,1536,0.0247715562582016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,1024,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,1536,0.004928888960017098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,1024,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,768,0.003903111235962974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,1024,0.023735110958417256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,768,0.023411555422676936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,768,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,512,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,512,0.022710222336981032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,512,0.011493333511882357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,256,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,256,0.02068444424205356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,128,0.0032479998966058097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,128,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,256,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,64,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3584,128,0.018256000346607633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,64,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3584,32,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,65536,0.07283555799060397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3584,32,0.010552888943089379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,51200,0.057322667704688184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,65536,0.1005057758755154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,65536,0.18545333544413248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,51200,0.08073511388566759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,16384,0.024846222665574815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,51200,0.1489244434568617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,16384,0.030574222405751545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,12288,0.019684443871180218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,16384,0.06380622254477607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,10240,0.016343111793200176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,12288,0.024600888291994732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,12288,0.051776889297697276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,8192,0.014105778601434497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,10240,0.04773511158095466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,10240,0.022376888328128394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,7168,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,8192,0.041533334387673274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,8192,0.020017777880032856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,6144,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,7168,0.018984888990720112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,7168,0.03978488842646281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,5120,0.010125333236323463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,6144,0.037704888317320086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,6144,0.017733333839310538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,5120,0.016499555773205228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,4096,0.00904444439543618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,5120,0.035637332333458796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,4096,0.014811555544535318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,3584,0.008603555460770925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,4096,0.03071999881002638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,3584,0.014422222971916199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,3584,0.031579554080963135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,3072,0.007704888780911763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,2560,0.0069857777820693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,3072,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,3072,0.02812622321976556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,2048,0.006080889039569431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,2560,0.027419555518362258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,2560,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,1536,0.0052826665341854095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,2048,0.013053333593739403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,2048,0.02531733281082577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,1024,0.004269333349333869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,1536,0.012531555361217923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,1024,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,768,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,1024,0.023617777559492324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,768,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,512,0.0034799998005231223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,768,0.022673777408070032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,512,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,256,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,512,0.021993777818149988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,256,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,128,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,256,0.0206995556751887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,1536,0.025426665941874187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,128,0.011071110765139261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,3072,128,0.019692444139056735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,64,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,3072,32,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,64,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,65536,0.06197333335876465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,3072,32,0.010820444259378644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,65536,0.0977413323190477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,51200,0.049729777706993945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,65536,0.18473954995473227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,51200,0.07849155531989203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,16384,0.02171733313136631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,51200,0.1490844488143921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,16384,0.06331200069851346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,12288,0.0177839994430542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,12288,0.024115555816226538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,12288,0.05177155468198988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,10240,0.014837332897716098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,16384,0.031123555368847314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,10240,0.022015111313925848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,10240,0.04796888762050205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,8192,0.01346933344999949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,8192,0.019996444384257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,8192,0.041489776637819074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,7168,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,7168,0.040151112609439425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,7168,0.018602665927675035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,6144,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,6144,0.03769333495034112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,6144,0.017844445175594754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,5120,0.010122666756312052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,4096,0.00886222223440806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,5120,0.03495733274353875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,5120,0.01591199967596266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,3584,0.008019555773999956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,4096,0.03228888909022013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,4096,0.014577777849303352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,3072,0.0074675555030504865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,3584,0.013904889424641928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,3584,0.029493331909179688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,2560,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,3072,0.013570666313171387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,3072,0.027238221632109746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,2560,0.013617777989970313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,2560,0.027862221002578735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,2048,0.013067555924256643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,1536,0.004575110971927643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,2048,0.025614221890767414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,1536,0.012492444780137805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,1536,0.024488000406159297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,1024,0.003951999876234266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,1024,0.012142221960756512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,768,0.003900444342030419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,2048,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,768,0.023023110296991136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,512,0.0035075553589397003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,512,0.011530666715568967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,1024,0.023736889163653057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,256,0.0032791110376516977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,512,0.022081777453422546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,256,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,256,0.0206675562593672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,128,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2560,128,0.019362666540675692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,64,0.0030373332815037835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,128,0.011030221978823343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2560,32,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,64,0.01167022188504537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,65536,0.05111022127999199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,32,0.01071200023094813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,65536,0.09517155753241645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,51200,0.041602667835023664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2560,768,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,65536,0.18493333127763537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,16384,0.01963555481698778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,51200,0.07609777980380587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,51200,0.14916443824768066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,12288,0.015413333972295126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,16384,0.02938400043381585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,16384,0.06356088982688056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,10240,0.014332445131407844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,12288,0.023561777340041265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,12288,0.05163288778728909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,8192,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,10240,0.04712000158098009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,10240,0.021720889541837905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,7168,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,8192,0.04181155562400818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,8192,0.019502222537994385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,6144,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,7168,0.018405333161354065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,7168,0.03936977850066291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,5120,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,6144,0.016549333930015564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,6144,0.03641066617435879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,4096,0.010070222119490305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,5120,0.015881778465376962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,5120,0.03379377722740173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,4096,0.030549334155188665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,3584,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,4096,0.014817777607176038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,3584,0.014154665999942355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,3072,0.008350222474998897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,3584,0.02949777907795376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,3072,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,3072,0.027843554814656574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,2560,0.007324444750944774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,2560,0.027523555689387854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,2560,0.013237333132161034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,2048,0.013058666553762225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,2048,0.026392888691690233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,1536,0.004954666727119022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,1536,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,2048,0.006247111078765657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,1536,0.025397333833906386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,1024,0.003994666867785984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,1024,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,768,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,768,0.01218933363755544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,768,0.023342221975326538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,512,0.00361333332127995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,512,0.011815110842386881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,1024,0.02443733314673106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,256,0.003168888803985384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,512,0.022343110707071092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,256,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,128,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,256,0.0218240006102456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,128,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,2048,128,0.019631110959582858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,32,0.002982222164670626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,64,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,2048,32,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,65536,0.04228533307711283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,65536,0.09207022190093994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,51200,0.034041778908835515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,65536,0.18492089377509224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,51200,0.07355111175113253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,16384,0.015443555182880826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,51200,0.15019999610053167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,2048,64,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,12288,0.01272266689274046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,16384,0.06328355603747897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,12288,0.023375110493765935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,10240,0.012764444781674279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,10240,0.0210479994614919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,12288,0.05198666784498426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,8192,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,10240,0.04696177774005466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,8192,0.018977777825461496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,7168,0.01125600023402108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,8192,0.040143112341562905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,16384,0.02885155545340644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,7168,0.01750577820671929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,6144,0.010409778190983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,6144,0.016026667422718473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,6144,0.03498488995763991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,5120,0.009385777844323052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,5120,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,5120,0.03324800067477756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,4096,0.008040000167157915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,4096,0.01442311041884952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,4096,0.02994133366478814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,3584,0.007076444725195567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,3584,0.013794666363133324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,3072,0.0063635553750726916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,3584,0.02953688965903388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,7168,0.03728355632887946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,2560,0.005612444546487596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,3072,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,3072,0.027782221635182697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,2048,0.005682666682534748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,2560,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,2048,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,2048,0.025436444414986506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,1536,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,1536,0.012427555190192329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,1536,0.02535466684235467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,1024,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,1024,0.02370400064521366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,768,0.004013333469629288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,2560,0.027161777019500732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,768,0.012058667010731168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,512,0.0033982222278912864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,768,0.021980444590250652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,512,0.011765333513418833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,512,0.021594666772418555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,256,0.003124444435040156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,1024,0.003863999827040566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,128,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,256,0.010755555497275459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,256,0.021874666213989258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,64,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,128,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1536,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1536,128,0.017663111289342243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,65536,0.030025776889589097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,32,0.011030221978823343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1536,64,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,65536,0.08955911133024429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,51200,0.025252444876564875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,65536,0.18540799617767334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,51200,0.150017778078715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,16384,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,16384,0.028920888900756836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,16384,0.06340533494949341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,51200,0.0704782207806905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,12288,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,12288,0.02254666719171736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,10240,0.01073688848151101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,12288,0.05182933476236132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,10240,0.01980533368057675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,10240,0.04731200138727824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,8192,0.009711110757456886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,8192,0.03996444410747952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,8192,0.018248889181349013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,7168,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,7168,0.03708533445994059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,7168,0.017542221479945712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,6144,0.01035199976629681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,6144,0.03463644451565213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,5120,0.009715555442704095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,6144,0.01622222198380364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,5120,0.014895111322402954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,5120,0.03285333183076646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,4096,0.0086986662613021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,4096,0.029967112673653498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,3584,0.00700711127784517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,4096,0.014434667097197639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,3584,0.013840888937314352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,3072,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,3584,0.02958044409751892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,3072,0.013402666482660504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,3072,0.027907556957668726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,2560,0.013311111264758639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,2560,0.02718488872051239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,2048,0.0058906666106647914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,2048,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,1536,0.0052693333062860705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,2048,0.025794666674402025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,2560,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,1024,0.004506666627195147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,1536,0.02475822303030226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,1536,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,1024,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,768,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,1024,0.023040889037979975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,512,0.0037626665499475268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,768,0.02306933369901445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,512,0.011518222590287527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,512,0.022704000274340313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,256,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,256,0.020385776956876118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,128,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,128,0.011101333631409539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,1024,128,0.01830222209294637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,768,0.011691555380821228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,64,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,1024,32,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,64,0.010853333605660332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,65536,0.025841777523358662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,32,0.010441777606805166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,1024,256,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,65536,0.08726666371027629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,51200,0.020873778396182593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,65536,0.18534133169386122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,16384,0.016116445263226826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,51200,0.06866311364703707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,51200,0.14953777525160047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,12288,0.012813333008024426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,16384,0.027808000644048054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,12288,0.05029333300060696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,12288,0.022069333328141108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,10240,0.0199608885579639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,10240,0.04542577928966946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,8192,0.010128888818952773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,16384,0.06183555391099718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,8192,0.03973155551486545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,8192,0.018375999397701688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,7168,0.009737778041097853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,10240,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,6144,0.00906488878859414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,7168,0.03706577751371596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,7168,0.01758666667673323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,5120,0.008716444174448649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,6144,0.03465777635574341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,6144,0.015863110621770222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,5120,0.03262933426433139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,4096,0.007378666765160031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,5120,0.015220445063379077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,3584,0.0070257774657673305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,4096,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,4096,0.0303182237678104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,3584,0.014170666535695394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,3584,0.029169778029123943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,3072,0.027790221903059218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,3072,0.013816888961527081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,2560,0.005903111149867375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,2560,0.013262222210566202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,2560,0.027220444546805486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,2048,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,3072,0.0064479998416370815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,2048,0.012823111481136747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,1536,0.0047333331571684945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,2048,0.026149332523345947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,1536,0.012510221865442065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,1536,0.024473778075642053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,1024,0.003989333493842019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,1024,0.02346666653951009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,1024,0.01185955521133211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,768,0.0036159998012913596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,768,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,768,0.02347022294998169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,512,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,512,0.011795555551846823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,512,0.021723555194007024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,256,0.0029884444342719186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,256,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,256,0.022037333912319605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,768,128,0.01831644442346361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,64,0.0029484445436133277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,768,32,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,64,0.01089866707722346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,65536,0.021512000097168818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,768,32,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,51200,0.017813333206706576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,65536,0.08525955677032471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,65536,0.18538667096032035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,16384,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,51200,0.06648355722427368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,51200,0.14827556080288357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,16384,0.0615288880136278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,16384,0.026467555099063452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,12288,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,10240,0.00978133330742518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,12288,0.021653332644038733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,12288,0.04972711205482483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,8192,0.009023110899660323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,10240,0.01960000064637926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,10240,0.044907556639777295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,7168,0.008041777544551426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,8192,0.01794933279355367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,8192,0.03910133242607117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,7168,0.016952888833151925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,6144,0.007592000067234039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,7168,0.03770844472779168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,6144,0.015850666496488783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,5120,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,6144,0.0343866679403517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,4096,0.007356444166766272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,5120,0.015024888846609326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,5120,0.0329751107427809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,3584,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,4096,0.030619555049472388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,4096,0.014172444740931192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,3072,0.0063564446237352155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,3584,0.014010666145218743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,3584,0.029581334855821397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,2560,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,3072,0.013163555827405719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,2560,0.013136000268989138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,2048,0.005140444470776452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,2560,0.02649333410792881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,2048,0.012759110993809171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,1536,0.004566222015354368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,2048,0.025819554924964905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,1536,0.012228444218635559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,1024,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,1536,0.0244159996509552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,1024,0.01187111106183794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,768,0.0038071109188927543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,3072,0.02757155564096239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,1024,0.023112888137499493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,512,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,768,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,768,0.023399111297395494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,512,0.022691556149058874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,256,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,512,0.011256888508796692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,128,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,256,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,256,0.021732444564501446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,512,128,0.018603555030292936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,512,32,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,128,0.011215111447705163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,65536,0.020426667398876615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,64,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,512,32,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,51200,0.01612711118327247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,65536,0.08171288834677802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,65536,0.18495378229353163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,51200,0.06572088930341932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,16384,0.009404444032245213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,16384,0.02611733310752445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,51200,0.14828444851769343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,12288,0.008353778057628209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,16384,0.06090488698747423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,12288,0.021659556362364028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,10240,0.0087013335691558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,12288,0.04933244321081373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,10240,0.019661333825853135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,10240,0.04539644387033251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,8192,0.01793066660563151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,8192,0.039804445372687444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,7168,0.007984889050324758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,7168,0.016911110944218107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,8192,0.00798044436507755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,7168,0.037669334146711565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,6144,0.007283555964628856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,6144,0.034507556094063654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,6144,0.015856888559129503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,5120,0.007634667058785756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,5120,0.03257066673702664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,5120,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,4096,0.007323555648326874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,4096,0.030527111556794908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,3584,0.006919999917348226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,4096,0.01443288889196184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,3584,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,3072,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,3584,0.029523554775449965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,3072,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,3072,0.02781244450145298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,2560,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,2560,0.027797334724002417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,2560,0.013167999684810638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,2048,0.005971555494599872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,2048,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,2048,0.02539377742343479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,1536,0.004320000194840961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,1536,0.012421333127551608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,1024,0.0035386664999855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,1536,0.025420443879233465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,1024,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,1024,0.02367288867632548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,768,0.0035075553589397003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,768,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,512,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,512,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,512,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,256,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,256,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,256,0.02202933364444309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,768,0.023021333747439917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,128,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,64,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,256,128,0.0192577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,256,32,0.0025155554629034465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,64,0.010392888552612728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,65536,0.01479644411140018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,32,0.011351111034552256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,65536,0.07991821898354425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,51200,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,65536,0.18482400311364067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,51200,0.14882755279541016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,16384,0.009363555245929295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,16384,0.02641777859793769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,16384,0.06052800019582113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,12288,0.007989333735571967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,51200,0.06329955657323201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,12288,0.02200711104604933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,12288,0.04913155568970574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,10240,0.007647111184067196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,256,128,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,10240,0.01923644377125634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,8192,0.006949333681000604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,10240,0.04517155554559496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,7168,0.006625777731339137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,8192,0.017258667283587985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,8192,0.03973155551486545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,6144,0.0063475556671619415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,7168,0.016950221525298226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,7168,0.03703555464744568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,5120,0.006004444426960415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,6144,0.015838222371207345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,6144,0.03536444571283128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,4096,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,5120,0.014447111222479077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,5120,0.03257600135273404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,3584,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,4096,0.014409777190950183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,4096,0.03012622065014309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,3072,0.00591466658645206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,3584,0.013739555246300168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,3584,0.029577778445349798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,2560,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,3072,0.013072888884279462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,3072,0.02811644474665324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,2048,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,2560,0.02713511056370205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,2560,0.013286222186353473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,1536,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,2048,0.012447111308574677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,2048,0.02476088868247138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,1024,0.0038364442686239877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,1536,0.02402044501569536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,1024,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,1024,0.02437422176202138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,768,0.0035031110876136353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,768,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,768,0.022732444935374792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,512,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,1536,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,512,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,256,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,512,0.022636445032225713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,256,0.011275555524561139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,128,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,256,0.02168088820245531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,128,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,64,0.0028115556471877625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,96,128,128,0.019343111250135634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,128,32,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,64,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,65536,0.012801777985360889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,65536,0.07923377884758843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,51200,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,51200,0.06257510847515531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,16384,0.008072000410821702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,16384,0.02608799934387207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,12288,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,12288,0.021599110629823472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,10240,0.0070222218831380205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,128,32,0.011179555621412067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,8192,0.0069546666410234236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,10240,0.019655999210145738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,7168,0.006676444576846228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,8192,0.018027555611398485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,6144,0.00664444433318244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,6144,0.0158906661801868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,5120,0.0063324446479479475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,4096,0.005439110928111606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,5120,0.015178667174445258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,3584,0.00600000015563435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,7168,0.016889777448442247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,3072,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,3584,0.01387644476360745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,4096,0.014010666145218743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,2560,0.006324444380071428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,2048,0.0058151111006736755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,3072,0.013460444079505073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,1536,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,2560,0.012834666503800286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,2048,0.012825777961148156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,1536,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,1024,0.0037066667444176147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,768,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,512,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,1024,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,256,0.002945777856641345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,512,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,768,0.011536888778209686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,256,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,128,0.010850666297806634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,64,0.0028204443968004654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,32,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,64,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,64,32,0.011398221883508893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,65536,0.011587555209795633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,51200,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,65536,0.07880622148513794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,16384,0.006966222491529252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,64,128,0.0026177778426143858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,12288,0.007281777759393056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,16384,0.02608977754910787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,12288,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,10240,0.01962222158908844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,8192,0.007002666592597961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,8192,0.017895999881956313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,7168,0.006654222392373615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,7168,0.016920000314712524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,51200,0.06083199712965223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,6144,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,5120,0.006261333409282897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,10240,0.0070159998204973005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,6144,0.01568800045384301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,5120,0.01497155593501197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,4096,0.005961777849329843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,3584,0.005964444329341252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,3072,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,4096,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,3584,0.013872000078360239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,3072,0.013423110875818463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,2048,0.006173333360089197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,2560,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,2048,0.012700444294346703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,1536,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,1024,0.0037804444630940757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,1536,0.012455999851226807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,768,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,1024,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,768,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,512,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,256,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,2560,0.005961777849329843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,256,0.011366222467687396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,128,0.002565333412753211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,128,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,64,0.002567111204067866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,64,0.010870222416188983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,512,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,96,32,32,0.0025066667132907445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,96,32,32,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,51200,1.0849822362263997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,51200,0.6210657755533854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,16384,0.36652088165283203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,16384,0.20760177241431343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,12288,0.2690009011162652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,12288,0.15823200013902453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,10240,0.21280977461073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,10240,0.13535822762383357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,8192,0.17608266406589082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,7168,0.14860444598727757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,8192,0.11209511756896973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,6144,0.133349339167277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,7168,0.10129955742094253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,5120,0.11153244972229004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,6144,0.08817955520417954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,5120,0.076401776737637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,4096,0.08902044428719415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,3584,0.07994133234024048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,4096,0.06429244412316217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,3072,0.06937511099709405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,2560,0.05832711193296644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,3072,0.05266311102443271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,2048,0.04997511042488945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,2560,0.04482577906714546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,3584,0.05859466393788656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,1536,0.037792887952592634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,1024,0.026483555634816486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,2048,0.03991377684805129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,768,0.020788444413079154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,1536,0.04703911145528158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,512,0.014514666464593677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,1024,0.03633244501219855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,768,0.029174221886528864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,256,0.009358222285906473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,128,0.00794044468137953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,256,0.01863822175396813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,512,0.023367111881573994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,64,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,65536,32,0.0059164443777667145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,128,0.016917333006858826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,64,0.01811377704143524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,65536,32,0.019828443725903828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,65536,0.6378764576382107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,51200,0.8299457761976453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,65536,1.0550337897406685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,16384,0.2822693453894721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,51200,0.4954835573832194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,16384,0.1679635577731662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,12288,0.20716799630059135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,12288,0.1280533340242174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,10240,0.17452711529201934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,10240,0.10966755284203424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,8192,0.08885688914193048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,7168,0.12141866154140896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,7168,0.08017422093285455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,6144,0.10707022084130181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,5120,0.08850666549470688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,6144,0.07083555724885729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,4096,0.07407377825842963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,5120,0.06132533152898153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,3584,0.06439733505249023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,4096,0.05148977703518338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,3072,0.056458665264977344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,3584,0.047071999973720975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,8192,0.14479555024041071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,2560,0.04759822289148966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,2048,0.04071999920739068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,3072,0.042860445049073964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,1536,0.030822220775816176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,2560,0.03828444414668613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,2048,0.03353155652681986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,1024,0.02243999971283807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,768,0.017217778497272067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,1536,0.038944890101750694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,1024,0.029890666405359905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,512,0.012711111042234631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,256,0.008994667066468133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,768,0.025791111919615004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,128,0.0058248887459437055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,512,0.020615999897321064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,256,0.01661066710948944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,64,0.004872888740566042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,51200,32,0.0052328887912962176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,64,0.01807822287082672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,32,0.017986666825082567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,51200,128,0.01589866644806332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,65536,0.3314017719692654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,51200,0.2579422261979845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,65536,0.253528012169732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,16384,0.09211555454466079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,12288,0.06664977471033733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,16384,0.07264888948864408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,51200,0.19819645086924234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,10240,0.056568887498643666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,12288,0.05576533079147339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,8192,0.04739200075467428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,10240,0.048650665415657886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,7168,0.04042933384577433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,8192,0.04120088948143853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,7168,0.037384000089433454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,6144,0.03618755605485704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,5120,0.030813333061006334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,6144,0.033788445923063494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,4096,0.0251208891471227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,5120,0.03048711021741231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,3584,0.022473777333895367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,4096,0.026147555973794725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,3072,0.019668444991111755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,3584,0.024634665913052026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,3072,0.023378666904237535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,2560,0.017047999633683097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,2048,0.01406044430202908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,2560,0.020564445190959506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,1536,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,2048,0.01904711127281189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,1024,0.008772444393899705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,1536,0.020437333318922255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,768,0.007749333149856991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,1024,0.01811377704143524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,512,0.005945777727497949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,768,0.0159262220064799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,256,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,512,0.014785778191354541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,256,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,128,0.003620444486538569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,64,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,128,0.012110222544935016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,16384,32,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,64,0.012815111213260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,16384,32,0.012793777717484368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,65536,0.2678479883405897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,65536,0.20092533694373238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,51200,0.20980355474683973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,16384,0.07518577575683594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,51200,0.1588195562362671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,12288,0.05473955472310384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,16384,0.060170663727654346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,10240,0.04688622223006355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,12288,0.04593066705597771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,8192,0.03923111160596212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,10240,0.04091733362939622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,7168,0.0337137778600057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,8192,0.03461155626508925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,6144,0.030105776256985132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,7168,0.03219288918707106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,6144,0.028774221738179524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,5120,0.025828444295459326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,4096,0.021290666527218286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,5120,0.025462221768167283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,3584,0.018921777606010437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,4096,0.02236266599761115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,3072,0.016536000702116225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,3584,0.0210479994614919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,2560,0.014579556054539151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,2560,0.01869422197341919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,3072,0.02001688877741496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,2048,0.012875555290116204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,1536,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,2048,0.017468444175190397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,1536,0.018435556027624343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,1024,0.008346666892369589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,768,0.007431999676757389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,1024,0.015584889385435315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,512,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,768,0.015180443723996481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,256,0.0041706665522522396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,512,0.01350044459104538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,128,0.003863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,256,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,128,0.01180355581972334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,64,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,12288,32,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,64,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,12288,32,0.013234666652149625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,65536,0.20444977283477783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,51200,0.15932710965474448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,65536,0.19457599851820204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,16384,0.06068978044721815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,51200,0.15374133321974012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,12288,0.04243466589185926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,16384,0.05825777848561605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,10240,0.03718577822049459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,12288,0.04436888959672716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,8192,0.03248000144958496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,10240,0.03956977857483758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,7168,0.027394667267799377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,8192,0.0336426662074195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,6144,0.02381244467364417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,7168,0.03126400046878391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,5120,0.020967110991477966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,6144,0.027219555444187585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,4096,0.017098666893111337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,5120,0.02440799938307868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,4096,0.021853332718213398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,3584,0.01555999947918786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,3584,0.02104088829623328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,3072,0.013937777943081327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,3072,0.01958666741847992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,2560,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,2560,0.01867377758026123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,2048,0.010432889064153036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,1536,0.00908799966176351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,1536,0.01794933279355367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,1024,0.007098666495747036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,2048,0.016992888516849942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,768,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,1024,0.01536266671286689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,512,0.00488800017370118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,768,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,256,0.003936000168323517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,512,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,128,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,256,0.012125333150227865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,128,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,64,0.003269333392381668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,10240,32,0.003301333429084884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,64,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,10240,32,0.012034667034943899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,65536,0.17276444700029162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,51200,0.13383732901679143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,65536,0.15171999401516384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,16384,0.0514337784714169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,51200,0.12042221758100723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,12288,0.0366924438211653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,16384,0.04740711053212484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,10240,0.031677332189348005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,12288,0.03677333394686381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,8192,0.02665688925319248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,10240,0.032800889677471586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,7168,0.023111111587948267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,8192,0.02777066661251916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,6144,0.02070399953259362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,7168,0.025272889269722834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,5120,0.017927999297777813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,6144,0.023071999351183575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,4096,0.015098666151364645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,5120,0.021039111746682063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,3584,0.013384000294738345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,4096,0.019609777463806998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,3072,0.011928888658682505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,3584,0.018599111172888014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,3072,0.01792711185084449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,2560,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,2048,0.00944800012641483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,1536,0.008064889245563084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,2560,0.0170160002178616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,1024,0.0058000000814596815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,2048,0.015626667274369132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,1536,0.015955555770132277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,768,0.004716444346639845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,512,0.004021333323584663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,1024,0.014740443891949125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,256,0.0036231109665499795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,768,0.013831111292044321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,512,0.012840888566441007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,128,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,64,0.003257777748836411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,256,0.01202400028705597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,128,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,64,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,8192,32,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,65536,0.15795111656188965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,51200,0.12143466207716201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,65536,0.14770755502912733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,16384,0.046462221278084644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,51200,0.11703288555145264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,16384,0.04620711008707682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,12288,0.03422755665249295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,10240,0.032636443773905434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,8192,32,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,12288,0.035437332259284124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,8192,0.02828977836502923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,7168,0.022082666556040447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,10240,0.03159200151761373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,6144,0.020630222227838304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,7168,0.024459555745124817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,5120,0.017435555656750996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,6144,0.023036445180575054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,4096,0.015509333875444202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,5120,0.02088266611099243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,3584,0.013968888256284924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,4096,0.01919555498494042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,3072,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,3584,0.018558222386572096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,2560,0.011205332974592844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,3072,0.017280000779363845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,8192,0.026712889472643535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,2048,0.009752888646390703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,2560,0.016767111089494493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,1536,0.007828444242477417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,1024,0.006068444500366847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,2048,0.01514844430817498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,1536,0.01612711118327247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,768,0.005305777821275923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,1024,0.014244443840450711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,512,0.00461777796347936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,768,0.013923555612564087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,256,0.003912888881233004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,512,0.012819555898507437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,128,0.0034151110384199354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,256,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,64,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,128,0.011172444456153445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,7168,32,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,64,0.011878222227096558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,65536,0.1274497773912218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,7168,32,0.011795555551846823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,51200,0.10140355428059895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,65536,0.14301599396599665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,16384,0.040138665172788836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,51200,0.11363022857242161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,12288,0.031154665682050917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,16384,0.044938666952980884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,10240,0.02793244520823161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,12288,0.03475111060672336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,8192,0.023238221804300945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,10240,0.03124177787039015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,7168,0.018266666266653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,8192,0.02608799934387207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,6144,0.01686044368478987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,7168,0.024123556084103052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,5120,0.014504000544548035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,6144,0.022084444761276245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,4096,0.012235555383894177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,5120,0.020690666304694284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,3584,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,4096,0.01886755559179518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,3072,0.011038222246699862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,3584,0.01831822262869941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,2560,0.009388444324334463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,3072,0.017324444320466783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,2048,0.008099555969238281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,2560,0.016007110476493835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,1536,0.007233777807818518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,2048,0.014160888062583076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,1024,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,1536,0.015499555402331881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,768,0.0046311111913787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,1024,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,768,0.013487111363146039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,512,0.004279110994603899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,256,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,512,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,128,0.003269333392381668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,256,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,64,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,128,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,6144,32,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,64,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,65536,0.11164622836642796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,6144,32,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,51200,0.08849600288603042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,16384,0.03577333357599046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,65536,0.1401804420683119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,51200,0.11166577868991429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,12288,0.028631998433007136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,16384,0.0432640016078949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,10240,0.025245333711306255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,8192,0.021311110920376245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,12288,0.03391911254988776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,7168,0.01885600056913164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,8192,0.02532088922129737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,6144,0.014472888575659858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,7168,0.023761777414215937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,5120,0.012892444100644855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,6144,0.021966222259733412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,4096,0.01074222226937612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,5120,0.02033333314789666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,3584,0.010536889235178629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,4096,0.018582221534517076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,3072,0.010418666733635796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,3584,0.018019555343521964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,2560,0.008968888885445064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,10240,0.028636445601781208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,2048,0.0074888889988263445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,3072,0.01627200014061398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,1536,0.006004444426960415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,2560,0.01514488955338796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,2048,0.013835555149449242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,1024,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,768,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,1536,0.014127110441525778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,1024,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,512,0.004315555509593752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,256,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,768,0.013131555583741931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,128,0.0031519999934567344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,512,0.012445333103338877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,64,0.0031475555151700974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,256,0.01184800018866857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,5120,32,0.0032524443748924467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,128,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,64,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,65536,0.09451733032862346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,5120,32,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,51200,0.07205688953399658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,65536,0.0995982223086887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,16384,0.03087200058831109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,51200,0.08032888836330838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,16384,0.03261422117551168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,12288,0.02466222147146861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,10240,0.021779555413458083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,12288,0.02491377790768941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,8192,0.015782222151756287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,10240,0.022721777359644573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,8192,0.020319110817379422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,7168,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,6144,0.012737777498033313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,7168,0.01897333396805657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,5120,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,6144,0.017980444762441847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,4096,0.009712888962692684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,5120,0.01719555589887831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,3584,0.0092275556590822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,4096,0.015871110889646742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,3584,0.014737778239780001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,3072,0.008438222110271454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,2560,0.007703999678293864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,3072,0.013760888742076026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,2048,0.006312888943486744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,2560,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,1536,0.004934222333961063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,2048,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,1024,0.004271999829345279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,1536,0.013532444834709167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,768,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,1024,0.012475555141766867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,512,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,768,0.01278222186697854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,512,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,256,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,128,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,128,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,64,0.003162666741344664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,64,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,32,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,4096,32,0.011173333558771344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,4096,256,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,65536,0.07914221949047513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,51200,0.06453777684105767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,65536,0.0974995560116238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,51200,0.07816977633370294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,16384,0.026744888888465032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,12288,0.022410665949185688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,16384,0.03065955638885498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,10240,0.01973155637582143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,12288,0.024367999699380662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,8192,0.017433777451515198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,10240,0.022107554806603327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,7168,0.013383111192120446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,7168,0.01867111192809211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,8192,0.019702222612169053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,6144,0.01218488895230823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,5120,0.010956444674068026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,6144,0.017575111654069688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,5120,0.016382222374280293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,4096,0.014835556348164877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,3584,0.008674666285514832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,3072,0.008115555677149031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,3584,0.014496889379289416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,2560,0.007146666447321574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,3072,0.01385244478782018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,2048,0.00565244464410676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,2560,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,1536,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,2048,0.013117333253224691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,1536,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,1024,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,768,0.003914666672547658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,4096,0.009121778110663096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,512,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,1024,0.012326222327020435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,256,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,768,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,512,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,256,0.011157333023018308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3584,32,0.003144888828198115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,128,0.011361777782440186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,64,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,65536,0.07045333253012763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3584,32,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,51200,0.05612711111704508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,16384,0.02402755618095398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,65536,0.09478844536675347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,12288,0.01886577738655938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,51200,0.0757404433356391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,10240,0.01664355562792884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,16384,0.029543999168607924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,8192,0.014727999766667684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,12288,0.02348622183005015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,10240,0.021300445000330608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,8192,0.019630221856964957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,7168,0.012561777399645912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,6144,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,7168,0.01847111185391744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,5120,0.010388444695207808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,6144,0.017403556240929496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,4096,0.008692444198661381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,5120,0.01587733295228746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,3584,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,4096,0.014152889450391134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,3072,0.007727999654081132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,3584,0.014146667387750415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,2560,0.0069520001610120135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,3072,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,2560,0.013223110801643796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,2048,0.005934222290913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,1536,0.004996444616052839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,2048,0.012770666844315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,1536,0.01273333364062839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,768,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,1024,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,512,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,768,0.012476444244384766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,512,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,128,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,256,0.011062222222487131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,64,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,128,0.011351111034552256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,3072,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,64,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,3072,32,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,65536,0.06025333536995781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,51200,0.04835022158092923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,65536,0.09165866507424249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,16384,0.02163644466135237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,51200,0.07312622335222033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,12288,0.01737244427204132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,16384,0.028836445675955877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,10240,0.015557333827018738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,12288,0.02309600015481313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,10240,0.021336000826623704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,8192,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,7168,0.012832889126406776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,8192,0.01924977699915568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,6144,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,7168,0.01789777808719211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,5120,0.010030222435792288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,6144,0.01699911057949066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,4096,0.008991999758614434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,5120,0.015464888678656684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,3584,0.007999111380841997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,4096,0.01444533301724328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,3072,0.007420444654093848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,3584,0.014056889547242058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,3072,0.013680000272062091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,2560,0.006550222221348021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,2048,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,2560,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,1536,0.004867555366622077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,2048,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,1024,0.003937777959638172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,1536,0.012827555338541666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,768,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,1024,0.012119111087587146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,512,0.0035866668654812705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,768,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,256,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,512,0.011824000212881299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,256,0.011527111132939657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,128,0.0031537777847713898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,64,0.0029173334025674393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,128,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2560,32,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,64,0.010793777803579966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,65536,0.04936977889802721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2560,32,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,51200,0.04219733344184028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,16384,0.01735111077626546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,65536,0.08861688772837321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,12288,0.015205333630243937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,16384,0.028146667612923518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,10240,0.013853333062595792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,12288,0.02275733318593767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,8192,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,10240,0.021088000800874498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,7168,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,8192,0.018613333503405254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,7168,0.017323555217848886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,6144,0.010823999842007955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,5120,0.009762666291660732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,6144,0.016208888755904306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,4096,0.008273777862389883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,5120,0.014910222755538093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,3584,0.007652444144090016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,4096,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,3072,0.00683111117945777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,3584,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,3072,0.013471999930010902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,2560,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,51200,0.0717359979947408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,2048,0.005241777747869492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,2560,0.013411555853154926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,1536,0.004596444467703502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,2048,0.012832000023788877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,1536,0.012529777983824411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,1024,0.004247111164861255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,768,0.003944000022278892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,1024,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,512,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,768,0.01221866657336553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,256,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,512,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,128,0.002982222164670626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,128,0.011101333631409539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,256,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,2048,32,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,32,0.011708444191349877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,65536,0.041229334142473005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,2048,64,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,51200,0.033055110110176936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,65536,0.08601244290669759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,16384,0.016567111015319824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,51200,0.06895555390252008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,16384,0.027809778849283855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,12288,0.022343110707071092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,10240,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,10240,0.020808888806237113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,8192,0.011203555597199334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,8192,0.01756177842617035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,7168,0.010777778095669217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,12288,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,7168,0.016937777400016785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,6144,0.010288000106811523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,5120,0.009309333231714036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,6144,0.015590222345458137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,4096,0.007669332954618666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,4096,0.014147554834683737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,5120,0.015218666858143277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,3584,0.00701422244310379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,3072,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,3072,0.013539555172125498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,3584,0.014134221606784396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,2560,0.005919111271699269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,2048,0.0053031109273433685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,2048,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,2560,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,1536,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,1024,0.003965333518054751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,1024,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,1536,0.012770666844315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,768,0.012148444851239523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,512,0.0033617777129014335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,512,0.012126222252845764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,256,0.0031537777847713898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,768,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,128,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,256,0.011533333195580376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,64,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,128,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1536,32,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,64,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1536,32,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,65536,0.029373334513770208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,51200,0.02475555572244856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,65536,0.0832666688495212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,16384,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,12288,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,51200,0.0660159985224406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,16384,0.02757599949836731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,12288,0.02165866725974613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,10240,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,8192,0.009769777456919352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,7168,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,10240,0.0192275560564465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,6144,0.010674667027261523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,7168,0.016512889001104567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,8192,0.018016000588734944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,5120,0.009664889011118147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,4096,0.008654221892356873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,6144,0.015508444772826301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,3584,0.0069671107663048645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,5120,0.015089778436554803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,4096,0.014152889450391134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,3072,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,3584,0.01384888920519087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,2560,0.00591555568906996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,3072,0.013456000222100152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,2048,0.006033777776691649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,1536,0.005355555564165115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,2560,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,1024,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,2048,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,768,0.003939555750952827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,1024,0.01182933317290412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,1536,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,512,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,768,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,256,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,256,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,512,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,128,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,64,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,64,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,128,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,1024,32,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,65536,0.0254737784465154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,1024,32,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,65536,0.08137510882483588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,51200,0.020815999971495736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,16384,0.01552888916598426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,51200,0.06344355477227105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,12288,0.013245333400037555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,16384,0.026495999760097925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,12288,0.021018667353524104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,10240,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,8192,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,10240,0.019281778070661757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,7168,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,8192,0.01757422255145179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,7168,0.016403555870056152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,6144,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,5120,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,6144,0.015582222077581616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,4096,0.007286222444640265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,5120,0.015091554986106025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,4096,0.014119999276267158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,3584,0.006622222148709827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,3072,0.006321777900060018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,3584,0.013824000126785703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,2560,0.00564177789621883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,3072,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,2048,0.005233777893914117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,2560,0.01315111087428199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,1536,0.004568888909286923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,1536,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,2048,0.012835555606418185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,1024,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,768,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,768,0.012167111039161682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,1024,0.012173333101802401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,512,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,256,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,512,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,256,0.011502222054534487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,128,0.002963555562827322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,64,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,128,0.011128889189826118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,768,32,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,32,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,65536,0.01831377711560991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,65536,0.07912444406085543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,51200,0.015259555644459195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,51200,0.061632891496022545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,16384,0.011790222591824003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,16384,0.0251217782497406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,12288,0.010067555639478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,768,64,0.011380444798204633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,10240,0.010301333334710863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,12288,0.02087644404835171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,8192,0.008743111458089616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,10240,0.019293333093325298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,7168,0.008385777473449707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,8192,0.01741155485312144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,6144,0.007376888559924231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,6144,0.015750222735934787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,5120,0.007976888782448238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,7168,0.016612443659040663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,4096,0.007313778003056844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,5120,0.014800889624489678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,3584,0.006691555596060223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,4096,0.014167999227841696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,3072,0.006333333336644703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,3584,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,2560,0.005519999812046687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,2560,0.01310755560795466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,3072,0.013510222236315409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,2048,0.004927111168702443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,1536,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,1536,0.012450666891203986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,1024,0.003896000070704354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,2048,0.012813333008024426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,768,0.003954666770166821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,1024,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,512,0.003551111039188173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,768,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,512,0.011824888487656912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,256,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,128,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,256,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,128,0.011448889142937131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,64,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,512,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,65536,0.01865600049495697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,512,32,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,51200,0.01587022178702884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,65536,0.07597511344485812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,51200,0.06089244286219279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,16384,0.011500444677140979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,16384,0.025507556067572698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,12288,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,12288,0.02113155523935954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,10240,0.009010666774378883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,10240,0.0192577772670322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,8192,0.008108444511890411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,8192,0.01755466726091173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,7168,0.008731555607583787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,7168,0.01660977800687154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,6144,0.00853777759604984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,5120,0.008533333738644918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,6144,0.015764445066452026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,4096,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,5120,0.014448000325096978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,3584,0.006748444504208035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,4096,0.014177777700954013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,3072,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,3584,0.01367911116944419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,2560,0.00572622236278322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,3072,0.013467555244763693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,2048,0.004987555659479565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,2560,0.012839111189047495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,1536,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,2048,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,1024,0.00387999994887246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,1536,0.012215110990736218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,768,0.003547555456558863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,1024,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,512,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,768,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,256,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,512,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,256,0.01111644423670239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,64,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,128,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,256,32,0.002576888849337896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,64,0.01070755554570092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,65536,0.014780445231331719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,256,32,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,51200,0.01440088947614034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,16384,0.009328000247478485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,65536,0.07465510898166232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,12288,0.009391110804345872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,51200,0.0592106646961636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,10240,0.008691555923885768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,16384,0.025058666865030926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,8192,0.008009778128729926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,12288,0.020668443706300523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,7168,0.00867733359336853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,10240,0.018815999229749043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,6144,0.00814844419558843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,8192,0.017462222112549674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,5120,0.007359999749395583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,7168,0.016208888755904306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,6144,0.015104000767072042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,4096,0.007033777733643849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,3584,0.006631999793979857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,4096,0.013793778088357715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,5120,0.015048889650238885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,3072,0.0062871111763848206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,2560,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,3584,0.01350400017367469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,2048,0.004899555610285865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,3072,0.01347644461525811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,2560,0.012832889126406776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,1536,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,1024,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,2048,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,1536,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,768,0.0032799999333090256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,512,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,1024,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,768,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,256,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,512,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,256,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,64,0.0025537777692079544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,128,32,0.0031191110610961914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,64,0.010473777850468954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,128,0.01149511088927587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,65536,0.015469332536061605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,128,32,0.010650667051474253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,51200,0.014496889379289416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,16384,0.009000000026490953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,65536,0.07396088706122504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,12288,0.009183111290136972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,12288,0.0209840006298489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,10240,0.008771555291281806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,10240,0.01938666734430525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,8192,0.008037333687146505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,8192,0.01751733322938283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,16384,0.02474044428931342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,7168,0.007680888805124495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,6144,0.007222221957312689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,7168,0.015885333220163982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,6144,0.01610133383009169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,5120,0.006711999989218182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,5120,0.014895111322402954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,4096,0.006664000037643645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,51200,0.05661244524849785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,3584,0.00600000015563435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,4096,0.014151111245155334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,3072,0.006653333289755716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,3584,0.01350844403107961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,2560,0.005978666659858491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,3072,0.01349511080318027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,2048,0.005962666538026597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,2560,0.012777778009573618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,1536,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,2048,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,1024,0.0037537775933742523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,1536,0.012264889147546558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,768,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,1024,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,512,0.003305777700410949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,768,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,256,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,512,0.011921777493423887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,256,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,128,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,64,32,0.002559111143151919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,64,0.010749333434634738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,65536,0.015497777197096082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,64,32,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,51200,0.01457244488928053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,65536,0.07156266768773396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,16384,0.00903911143541336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,51200,0.0565155545870463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,16384,0.024824000067181055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,12288,0.00963466697269016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,10240,0.008711111214425828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,12288,0.020583111378881667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,8192,0.007672000262472365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,10240,0.0192284451590644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,7168,0.007318221860461765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,8192,0.01719288859102461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,6144,0.006659555352396435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,7168,0.016534222496880423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,5120,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,6144,0.015406222807036506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,4096,0.006491555521885554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,3584,0.00646311127477222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,4096,0.013836444252067141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,3584,0.013925333817799887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,3072,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,2560,0.005960888746711943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,3072,0.013451555536852943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,2048,0.00564533347884814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,2048,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,2560,0.012803555362754397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,1536,0.004231111043029361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,1024,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,1024,0.01180622229973475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,5120,0.014888889259762235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,768,0.003588444242874781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,1536,0.01222222215599484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,512,0.00332622230052948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,256,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,768,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,256,0.011192888849311404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,128,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,512,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,64,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,81,32,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,128,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,64,0.010775110787815519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,81,32,32,0.011530666715568967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,51200,0.6210248735215929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,51200,1.0530435774061415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,16384,0.3558355437384711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,16384,0.20728088749779594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,12288,0.2641688982645671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,12288,0.1580266687605116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,16384,0.2623822159237332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,12288,0.20324356026119658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,10240,0.2133715550104777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,10240,0.13511644469367132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,8192,0.17071733209821913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,8192,0.10892355442047119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,10240,0.1739955610699124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,7168,0.14967111746470133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,8192,0.14545688364240858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,7168,0.09788978099822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,51200,0.7569813198513455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,6144,0.13087911076015896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,6144,0.08673689100477432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,7168,0.12991555531819662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,5120,0.10803822014066909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,6144,0.11631645096672906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,5120,0.07421955797407362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,4096,0.08804355727301703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,4096,0.061906668874952525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,3584,0.07748088571760389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,5120,0.10246844424141778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,4096,0.08798133002387153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,3584,0.05573778020011055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,3072,0.06705955664316814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,3584,0.07976977692710029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,3072,0.07375644312964545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,2560,0.058099554644690625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,3072,0.05076711045371162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,2560,0.04432622260517544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,2048,0.0476817786693573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,2560,0.0667182207107544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,2048,0.03833955526351929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,1536,0.03701600101259019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,2048,0.059544000360700816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,1536,0.042191998826132886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,1024,0.026421333352724712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,1536,0.061158219973246254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,1024,0.030101332399580214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,1024,0.05014666583802965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,768,0.020809777908855014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,768,0.02827111217710707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,768,0.040806220637427434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,512,0.014476444986131458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,512,0.03713422351413303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,512,0.02199999988079071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,256,0.009095999929640029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,256,0.017877333694034152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,256,0.03362222181426154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,128,0.007282666862010956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,128,0.01726222203837501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,65536,128,0.026386666629049513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,64,0.0063271112740039825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,65536,32,0.006674666785531574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,64,0.016859556237856548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,65536,32,0.021724444296624925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,65536,0.6370906829833984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,65536,1.0524177551269531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,51200,0.828006214565701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,51200,0.4951324462890625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,16384,0.2836577892303467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,16384,0.1677751143773397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,16384,0.22876622941758898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,12288,0.2066648933622572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,12288,0.12767822212643093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,65536,0.8337359958224826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,10240,0.17462666829427084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,10240,0.10987377829021877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,51200,0.6567431026034886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,8192,0.1446711089875963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,12288,0.17782399389478895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,8192,0.08903377585940891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,10240,0.15287110540601942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,7168,0.1213031080034044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,8192,0.1275884442859226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,7168,0.08027911186218262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,6144,0.10628977749082778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,7168,0.11466044849819607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,5120,0.08881333139207627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,6144,0.07102577553855048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,6144,0.10312977764341567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,5120,0.06133244434992472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,4096,0.07434755563735962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,5120,0.09138933155271743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,4096,0.05176888902982076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,3584,0.06387555599212646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,3072,0.055572443538241915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,3584,0.046931554873784385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,3584,0.07252622312969632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,3072,0.04314755400021871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,2560,0.04712533288531833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,3072,0.06608977582719591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,2048,0.04012088974316915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,2560,0.037976887491014265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,2560,0.06092800034417046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,2048,0.0333555539449056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,1536,0.030396444929970637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,2048,0.054449776808420815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,1024,0.021926222576035395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,4096,0.07870666848288642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,1536,0.0542702211274041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,1536,0.035585777627097234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,768,0.016568000117937725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,1024,0.025536000728607178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,512,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,1024,0.04335911075274149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,768,0.03864088985655043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,768,0.021920000513394673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,256,0.008235555556085374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,512,0.034019556310441755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,512,0.020236444142129686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,128,0.0053431110249625305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,256,0.030135111676322088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,256,0.016180444094869826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,64,0.004624888714816836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,51200,32,0.004599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,128,0.01585422290696038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,51200,128,0.0239982224173016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,64,0.015486222174432544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,51200,32,0.01926577753490872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,65536,0.32278667555914986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,51200,0.2500080002678765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,65536,0.2518124315473768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,51200,0.1981431113349067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,16384,0.08839378092024062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,16384,0.07257066832648383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,16384,0.08157955275641547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,12288,0.06418310933642916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,65536,0.24552800920274523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,12288,0.055514666769239635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,10240,0.05469510952631632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,12288,0.06564266814125909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,10240,0.04876266585456001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,8192,0.0454346670044793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,8192,0.040821333726247154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,10240,0.058020446035597056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,7168,0.039911998642815485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,51200,0.19364444414774576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,8192,0.05151822169621786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,6144,0.034789333740870156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,7168,0.037717332442601524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,7168,0.047204444805781044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,6144,0.033815112378862165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,6144,0.044806222120920815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,5120,0.029778665966457788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,4096,0.024135111106766596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,5120,0.03046311272515191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,5120,0.040567109982172646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,4096,0.02621600031852722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,3584,0.022134222918086584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,4096,0.0370506677362654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,3072,0.0195768889453676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,3584,0.023644444015291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,3584,0.03489689032236735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,2560,0.01650399963061015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,3072,0.02239911092652215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,3072,0.03438933359252082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,2048,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,2560,0.032638221979141235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,2560,0.02056711084312863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,2048,0.03120622369978163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,1536,0.010875555376211802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,2048,0.01882577770286136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,1024,0.00868800034125646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,1536,0.019537778364287484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,1536,0.028899557060665552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,1024,0.01583377851380242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,768,0.007227555745177799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,1024,0.02773688899146186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,512,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,768,0.01479288935661316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,768,0.02511288887924618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,256,0.0037884443170494507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,512,0.013768889009952545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,512,0.022637334134843614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,128,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,256,0.020959999826219346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,128,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,256,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,64,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,16384,128,0.01957244508796268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,16384,32,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,64,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,16384,32,0.014137778017255994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,65536,0.2579102251264784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,65536,0.20151289304097494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,51200,0.19762489530775282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,51200,0.15838578012254503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,16384,0.07524710893630981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,16384,0.05987199809816149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,65536,0.21580889489915636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,12288,0.05461155705981784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,51200,0.17055288950602213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,12288,0.04609333475430807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,10240,0.0462719996770223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,16384,0.07399288813273112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,12288,0.06013244390487671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,8192,0.0391813317934672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,10240,0.040736890501446195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,10240,0.05294488867123922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,7168,0.03388444582621256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,8192,0.034648888640933566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,8192,0.047622223695119224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,6144,0.029851555824279785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,7168,0.04381510946485731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,6144,0.02903999884923299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,7168,0.032123555739720665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,5120,0.025895110434956018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,6144,0.041095998552110456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,4096,0.021320889393488567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,5120,0.03836266530884637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,4096,0.022242666946517095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,3584,0.019144889381196763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,3584,0.0206995556751887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,3584,0.033626665671666466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,3072,0.016307555966907077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,3072,0.02013777858681149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,3072,0.03259911139806112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,5120,0.025313778056038752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,2560,0.014263111684057446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,4096,0.03498222099410163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,2560,0.03124622172779507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,2048,0.01257155587275823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,2560,0.01869066721863217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,2048,0.017258667283587985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,1536,0.010116444693671333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,2048,0.029662221670150757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,1536,0.017064889272054035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,1024,0.008405333591832055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,1536,0.027979556057188246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,1024,0.014492443866199918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,768,0.007229333122571309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,1024,0.02635555631584591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,768,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,512,0.004972444640265571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,768,0.0236799998415841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,512,0.013156444662147097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,256,0.003948444293604957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,512,0.02236266599761115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,256,0.01180266671710544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,128,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,256,0.021312889125612047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,128,0.011575111084514193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,64,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,12288,128,0.01924088928434584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,12288,32,0.0035964445107513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,64,0.012379555238617791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,12288,32,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,65536,0.2041715515984429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,65536,0.192175997628106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,51200,0.1594337754779392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,51200,0.15327289369371203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,65536,0.20390310552385119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,16384,0.058244446913401283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,16384,0.057733330461714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,12288,0.04374400112364027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,16384,0.07093689176771376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,51200,0.16253777345021567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,10240,0.036453333165910505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,12288,0.044769777192009814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,12288,0.057504885726504855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,8192,0.03200088938077291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,10240,0.039503110779656306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,8192,0.03364177876048618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,10240,0.05220266514354282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,7168,0.028367112080256145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,8192,0.04559288753403557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,7168,0.04232622186342875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,6144,0.023846222294701472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,7168,0.03123377760251363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,5120,0.021219554874632094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,6144,0.027473777532577515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,6144,0.04002933369742499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,4096,0.017941332525677152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,5120,0.02464444438616435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,5120,0.036666668123669095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,3584,0.01609688831700219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,4096,0.035056889057159424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,4096,0.021539555655585393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,3584,0.020446222689416673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,3584,0.03287200133005778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,3072,0.019300444258583915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,2560,0.012664000193277994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,3072,0.032630221711264715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,2560,0.018374222848150466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,2048,0.011188444164064197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,2560,0.03073422114054362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,2048,0.016982222596804302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,1536,0.009321778184837764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,2048,0.029778665966457788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,1536,0.016172443826993305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,3072,0.014143110977278816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,1536,0.028186665640936956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,1024,0.014168888330459595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,768,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,1024,0.02511022157139248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,768,0.024648000796635944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,768,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,512,0.004967999955018361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,512,0.02309333284695943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,512,0.012870222330093384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,256,0.004219555606444676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,1024,0.007699555820888943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,128,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,256,0.022657778528001573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,128,0.011230222052998014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,64,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,10240,128,0.0199991116921107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,10240,32,0.0033653332955307434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,64,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,65536,0.17355822192298043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,32,0.013136000268989138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,65536,0.15082844098409018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,51200,0.1337653398513794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,10240,256,0.012085333466529846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,51200,0.11920800473954941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,16384,0.05095377895567152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,65536,0.18385067250993517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,16384,0.04709688822428385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,12288,0.036841776635911726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,16384,0.06449955701828003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,51200,0.14738578266567653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,12288,0.0365920000606113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,10240,0.03176000052028232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,12288,0.05366311139530606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,10240,0.03269511130121019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,8192,0.02667288813326094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,10240,0.048116445541381836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,8192,0.028152889675564233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,7168,0.02316088808907403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,8192,0.04317422376738655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,7168,0.024804444776640996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,7168,0.04080977704789903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,6144,0.020671111014154222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,6144,0.03851111067665948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,5120,0.01795022189617157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,5120,0.02107288936773936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,5120,0.03572266631656223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,4096,0.015124445160230001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,4096,0.01925066610177358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,4096,0.03382044368320041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,3584,0.01314133322901196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,6144,0.022879999544885423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,3584,0.018573333819707234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,3072,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,3584,0.03161244591077169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,3072,0.017476444443066914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,2560,0.010671111444632212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,3072,0.031253334548738264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,2560,0.016549333930015564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,2048,0.00941244430012173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,2560,0.029827555020650227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,2048,0.015455111861228943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,1536,0.007948444121413762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,2048,0.028532445430755615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,1024,0.006504888749784893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,1536,0.02681422233581543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,1024,0.01384888920519087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,768,0.00481244424978892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,1024,0.02403822210099962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,768,0.013350221845838757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,768,0.023066666391160753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,512,0.004266666869322459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,512,0.02274399995803833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,512,0.012409778104888068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,256,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,1536,0.01501511037349701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,128,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,256,0.021039111746682063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,256,0.011644444531864591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,64,0.003264888914095031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,128,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,8192,128,0.018918222851223417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,8192,32,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,64,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,65536,0.16198310587141249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,65536,0.14635554949442545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,51200,0.1235742171605428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,65536,0.18129066626230875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,51200,0.11597332689497207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,16384,0.047395557165145874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,8192,32,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,16384,0.06337066491444905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,16384,0.045907557010650635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,12288,0.03760622276200189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,51200,0.14636711279551187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,10240,0.03455111053254869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,12288,0.05296088920699226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,10240,0.031382222970326744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,8192,0.02587466604179806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,10240,0.04734399914741516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,8192,0.02658933401107788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,8192,0.04317155480384827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,7168,0.025071110990312364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,7168,0.024738666084077623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,7168,0.03967466619279649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,6144,0.021470222208235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,6144,0.022617777188618977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,6144,0.03804888990190294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,12288,0.03538666831122504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,5120,0.020610666937298246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,4096,0.014101333088344999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,5120,0.03535111082924737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,4096,0.018901333212852478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,3584,0.012810666528013019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,4096,0.03327466713057624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,3584,0.01829866733815935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,3072,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,3072,0.01756000022093455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,5120,0.017824888229370117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,2560,0.010391111175219217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,3072,0.03180711136923896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,2560,0.016183111402723525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,2560,0.030557334423065186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,2048,0.009105777574910058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,2048,0.015449777245521545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,2048,0.026791110634803772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,1536,0.007617777420414819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,1536,0.014548444085650973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,1536,0.02572533322705163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,1024,0.005632888939645555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,1024,0.013745778136783175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,1024,0.02395733363098568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,768,0.004830222162935468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,768,0.013080889152155982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,768,0.0236817780468199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,512,0.0038693332009845306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,3584,0.03151111139191522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,512,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,256,0.0033280000918441345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,512,0.02181688944498698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,256,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,128,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,128,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,256,0.020711110697852243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,64,0.00328533330725299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,7168,128,0.018926221463415358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,7168,32,0.0033280000918441345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,64,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,65536,0.12721333238813612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,65536,0.14022310574849448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,7168,32,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,51200,0.10103378030988906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,65536,0.17889955308702257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,51200,0.11309422387017144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,16384,0.04075555668936835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,16384,0.04471288786994087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,51200,0.1448097758822971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,12288,0.03128266665670607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,16384,0.06272711356480916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,12288,0.03426933288574219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,12288,0.05253511004977756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,10240,0.028780443800820246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,10240,0.030213332838482324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,10240,0.046837333175871104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,8192,0.02090399960676829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,8192,0.042415112257003784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,7168,0.018423999349276226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,8192,0.02539377742343479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,7168,0.023984000086784363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,6144,0.016226665841208566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,7168,0.039447112215889826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,6144,0.022406222091780767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,6144,0.03795911206139459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,5120,0.03494844502872891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,4096,0.012793777717484368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,5120,0.02090311050415039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,4096,0.018948444061809115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,3584,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,3584,0.017821333474583097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,4096,0.03390577766630385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,5120,0.014442667365074158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,3072,0.010996444357766045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,3584,0.03194400005870395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,2560,0.00960355583164427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,3072,0.016903110676341586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,3072,0.03170222375128005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,2560,0.030249777767393324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,2048,0.008059555457697975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,2560,0.01631644368171692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,1536,0.0069306666652361555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,2048,0.014736000034544202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,2048,0.02676800058947669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,1536,0.014135110709402295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,1024,0.005686222265164058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,1536,0.025786666406525507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,768,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,1024,0.02478133307562934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,768,0.012830221818553077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,512,0.0041955556306574075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,768,0.023769777682092454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,512,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,256,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,512,0.02312711046801673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,256,0.011815110842386881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,256,0.02204888893498315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,128,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,64,0.003023111157947116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,1024,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,6144,128,0.019655999210145738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,64,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,65536,0.11040355761845906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,6144,32,0.012530666258600024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,65536,0.13934755325317383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,51200,0.08983200126224095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,51200,0.11158755090501572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,65536,0.17801333798302543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,16384,0.0353724459807078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,6144,32,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,16384,0.06240799691942003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,16384,0.043861332866880626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,12288,0.02812444501452976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,12288,0.03287911083963182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,12288,0.051817778084013194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,10240,0.028772443532943726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,10240,0.04611644479963514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,8192,0.02150933278931512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,51200,0.14359645048777261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,8192,0.025026667449209426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,8192,0.04164355662133958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,10240,0.025099555651346844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,7168,0.023623999622133043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,7168,0.03964799973699782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,6144,0.014560000763999091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,6144,0.022320000661744013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,7168,0.018796443939208984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,5120,0.013000888956917657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,6144,0.03801155421468947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,5120,0.01995644470055898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,5120,0.035487999518712364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,4096,0.010826666322019366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,4096,0.018971555762820773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,3584,0.01012711144155926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,4096,0.033168001307381526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,3584,0.018016000588734944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,3584,0.03151911165979173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,3072,0.01126488877667321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,3072,0.016237333416938782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,3072,0.031269331773122154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,2560,0.009418666362762451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,2048,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,2560,0.015136000182893542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,2560,0.029002666473388672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,1536,0.006340444501903322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,2048,0.013830222189426422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,2048,0.026185777452256944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,1024,0.005223111146026188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,1536,0.025011556016074285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,1536,0.01388533330625958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,768,0.004917333523432414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,1024,0.023046221998002794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,1024,0.013148444394270578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,512,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,768,0.012510221865442065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,768,0.02256177696916792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,256,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,512,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,512,0.022143999735514324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,256,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,128,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,256,0.02200800014866723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,128,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,64,0.0032479998966058097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,5120,32,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,5120,128,0.017955554856194388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,64,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,65536,0.09382666481865777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,5120,32,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,51200,0.07109333409203424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,65536,0.09899644719229804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,51200,0.07842577828301324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,65536,0.17623644404941133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,16384,0.028005331754684448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,51200,0.14336444271935359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,12288,0.024698666400379602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,16384,0.03151466780238681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,16384,0.06272711356480916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,10240,0.021747555997636583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,12288,0.025813332862324182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,12288,0.051446222596698336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,10240,0.045624001158608325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,8192,0.015684444043371413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,10240,0.02312177750799391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,7168,0.014148443937301636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,8192,0.019984000258975558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,8192,0.04123377799987793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,6144,0.012726222475369772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,7168,0.038972444004482694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,7168,0.019389332996474374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,5120,0.011374221907721626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,6144,0.03717599974738227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,5120,0.034309334225124784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,5120,0.017179555363125272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,4096,0.009687999884287516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,4096,0.03286133209864298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,3584,0.009311111436949836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,3584,0.014526221487257214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,3584,0.03104089034928216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,3072,0.0083155557513237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,6144,0.018277333842383493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,3072,0.014519110321998596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,2560,0.007700444095664554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,3072,0.029341333442264136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,2560,0.013384888569513956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,2048,0.006665777828958299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,2560,0.02735822233888838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,4096,0.01593777702914344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,1536,0.005000000198682149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,2048,0.012832000023788877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,2048,0.026202667090627883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,1024,0.004253333227501975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,1536,0.025143999192449782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,1536,0.013206221991115145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,768,0.004022222426202563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,1024,0.022451554735501606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,1024,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,512,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,768,0.02274399995803833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,768,0.012451555165979596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,256,0.0033004445334275565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,512,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,512,0.021687999367713928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,256,0.02000177734427982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,256,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,128,0.010782221953074137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,4096,128,0.018019555343521964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,64,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,64,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,32,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,4096,128,0.002938666691382726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,4096,32,0.011217777927716574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,65536,0.07736621962653266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,65536,0.09646577967537774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,51200,0.06389511293835111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,65536,0.17635288503434923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,51200,0.07730044258965386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,16384,0.026154667139053345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,51200,0.14288000265757242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,16384,0.030039111773173015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,12288,0.02092355489730835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,16384,0.06192533175150553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,12288,0.023842665884229872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,12288,0.051539556847678296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,10240,0.019305777218606737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,10240,0.02204711072974735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,10240,0.04558755622969734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,8192,0.016507556041081745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,8192,0.041490667396121554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,7168,0.01739911072784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,8192,0.019688889384269714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,7168,0.018762666318151686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,6144,0.014144889182514615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,7168,0.0387368897596995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,6144,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,5120,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,6144,0.03724266754256354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,5120,0.016345777445369296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,4096,0.009380444056457942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,5120,0.03429777754677667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,4096,0.015240000353919135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,3584,0.008671110702885522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,4096,0.033004442850748696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,3584,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,3584,0.02948444419436985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,3072,0.0138586668504609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,3072,0.02852533260981242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,2560,0.006992888947327931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,2560,0.013184000220563678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,2048,0.006231110956933763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,3072,0.007827555967701806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,2560,0.027522666586769953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,2048,0.012789333032237159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,1536,0.005129777722888523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,2048,0.025794666674402025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,1024,0.004263999975389904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,1536,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,1536,0.025440000825458105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,768,0.003936888857020272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,1024,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,1024,0.023011555274327595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,512,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,768,0.012144889268610211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,768,0.022445332672860887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,256,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,512,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,512,0.02238133384121789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,256,0.010812444819344414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,128,0.0032053333189752367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,256,0.020000888241661918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,64,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,128,0.010710222025712332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3584,128,0.018927110566033255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3584,32,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,65536,0.0696835584110684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,64,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3584,32,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,51200,0.0547271105978224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,65536,0.09194933043585883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,51200,0.07365422116385566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,16384,0.02381155557102627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,65536,0.17575289143456352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,51200,0.14278666178385416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,16384,0.02920088834232754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,12288,0.018911111685964797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,12288,0.023747555083698694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,12288,0.051447110043631665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,16384,0.061194664902157254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,10240,0.016709332664807636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,8192,0.014779556128713818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,10240,0.021408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,8192,0.019388443893856473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,7168,0.012621333201726278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,8192,0.04152977797720168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,7168,0.018357333209779527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,7168,0.03863822089301215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,6144,0.011376000112957425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,6144,0.017186666528383892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,6144,0.03712977634535895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,5120,0.010676444404655032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,5120,0.03424533208211263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,5120,0.01610577768749661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,4096,0.009009777671760982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,4096,0.014858666393491956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,10240,0.04566399918662178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,4096,0.032272001107533775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,3584,0.008591111335489485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,3584,0.014301333162519666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,3072,0.007624000310897827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,3584,0.03125777840614319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,3072,0.013749333719412485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,2560,0.007241778075695038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,3072,0.029253333806991577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,2560,0.01314133322901196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,2048,0.006346666564544042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,2048,0.01276000009642707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,2560,0.02731200059254964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,1536,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,2048,0.026100445124838088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,1536,0.025128000312381323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,1024,0.004239999999602635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,1536,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,1024,0.012352888782819113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,768,0.0038782221575578055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,1024,0.02309866746266683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,768,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,512,0.0037502224246660867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,768,0.022487999664412603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,512,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,512,0.02234933276971181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,256,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,128,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,256,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,256,0.021899556120236714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,64,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,128,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,3072,128,0.01902577777703603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,3072,32,0.003120888852410846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,65536,0.059993777010175914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,32,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,3072,64,0.011621333658695221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,51200,0.04690222276581658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,65536,0.09088889095518325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,51200,0.07070666551589966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,65536,0.17568533950381807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,16384,0.021357332666714985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,16384,0.02865333358446757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,51200,0.14246043894026014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,12288,0.017289777596791584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,16384,0.06199200285805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,10240,0.015459555718633862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,12288,0.0509475568930308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,12288,0.0233271105421914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,8192,0.013835555149449242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,10240,0.045662220981385976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,8192,0.018972444865438674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,8192,0.041215999258889094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,7168,0.012880889077981314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,7168,0.018264888061417472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,6144,0.011236444115638733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,7168,0.038571556409200035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,6144,0.016604445046848722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,6144,0.036900444163216486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,10240,0.021398221453030903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,5120,0.016007999579111736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,4096,0.00888533310757743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,5120,0.03390844331847297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,4096,0.014485332700941296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,3584,0.008249777886602614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,4096,0.0317555566628774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,3584,0.013732444081041547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,3584,0.028816001282797918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,5120,0.009858667022652095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,3072,0.0074106670088238185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,2560,0.006603555546866522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,3072,0.028880000114440918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,3072,0.013767999907334646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,2048,0.0057839999596277876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,2560,0.013095999757448832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,2560,0.027316444449954565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,1536,0.00461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,2048,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,2048,0.025822222232818604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,1536,0.012653333445390066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,1024,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,1536,0.024752888414594863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,768,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,1024,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,1024,0.022713777091768052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,512,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,768,0.023360000716315374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,768,0.012089778151777057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,256,0.0032542221662071017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,512,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,512,0.022280888424979314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,128,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,256,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,256,0.02032444377740224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,128,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,64,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2560,32,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2560,128,0.019561777512232464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,64,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,65536,0.048199110560946994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2560,32,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,65536,0.0868426693810357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,51200,0.0403146677547031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,65536,0.17644445101420084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,51200,0.0707004467646281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,16384,0.01759288873937395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,16384,0.028173334068722192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,51200,0.14320533805423313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,12288,0.015182221929232279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,16384,0.06117244561513265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,12288,0.023013333479563396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,12288,0.05077777637375725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,10240,0.0141004439857271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,10240,0.021000888612535264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,8192,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,10240,0.045951998896069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,8192,0.019009777241282992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,7168,0.01146044416560067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,8192,0.041617777612474226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,7168,0.017550221747822232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,7168,0.039247112141715154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,6144,0.011012444065676795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,6144,0.03674577673276266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,6144,0.0165964447789722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,5120,0.010053333308961656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,5120,0.015522667103343539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,5120,0.03255911005867852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,4096,0.00996977753109402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,4096,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,3584,0.010062221851613786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,4096,0.03086400032043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,3584,0.013841778039932251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,3072,0.007975111405054728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,3072,0.013428444663683573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,3072,0.028211557202868994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,2560,0.008058666355080074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,2560,0.01310400002532535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,2560,0.027464888162083093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,2048,0.0052382221652401825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,3584,0.0296053323480818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,2048,0.012793777717484368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,1536,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,2048,0.025815111067559984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,1536,0.01240533341964086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,1024,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,1536,0.024792888098292883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,1024,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,768,0.003850666599141227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,1024,0.0239964442120658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,768,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,768,0.023383999864260357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,512,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,512,0.022295110755496558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,256,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,256,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,256,0.02064622276359134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,128,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,128,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,2048,128,0.018918222851223417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,64,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,2048,32,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,512,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,65536,0.04047733214166429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,64,0.010967999696731567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,65536,0.0852782196468777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,65536,0.17567111386193168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,51200,0.03348799877696567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,51200,0.06857599814732869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,16384,0.015491555134455362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,51200,0.1437351173824734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,16384,0.027217777238951787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,12288,0.012819555898507437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,16384,0.061511112584008105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,2048,32,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,10240,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,12288,0.022525333695941504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,12288,0.051157332128948636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,10240,0.04538844360245598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,8192,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,8192,0.018547554810841877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,8192,0.04162311222818162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,7168,0.010959111154079437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,7168,0.03821422325240241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,7168,0.017176000608338248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,6144,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,6144,0.03461155626508925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,5120,0.00942133367061615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,5120,0.01477866702609592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,10240,0.020408888657887776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,5120,0.03156088789304098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,4096,0.00794488853878445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,4096,0.013920888304710388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,4096,0.030208888981077407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,3584,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,3584,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,3584,0.028804444604449805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,3072,0.006949333681000604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,3072,0.013266666895813413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,6144,0.016177778442700703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,2560,0.006057777752478917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,3072,0.028208000792397395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,2560,0.02645422187116411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,2048,0.0052933332820733385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,2560,0.013236444857385425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,2048,0.012463111016485425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,1536,0.0046311111913787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,2048,0.025938666529125635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,1536,0.012152888708644442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,1024,0.0041706665522522396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,1024,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,1536,0.024688889582951862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,768,0.004105777790149053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,1024,0.022312889496485393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,768,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,768,0.02235466738541921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,512,0.003664888855483797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,512,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,256,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,512,0.02236888971593645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,256,0.010789333118332757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,256,0.02162577708562215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,128,0.0031724443866146933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1536,128,0.017613333132531907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,64,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,64,0.010755555497275459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1536,32,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,32,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,65536,0.027853333287768896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,65536,0.0803297758102417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1536,128,0.011599111060301462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,51200,0.023735110958417256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,65536,0.17660710546705458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,51200,0.14331555366516113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,51200,0.06360889143413968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,16384,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,16384,0.06091022491455078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,12288,0.011359111302428775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,16384,0.02790133489502801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,10240,0.010814222196737925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,12288,0.021706667211320665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,12288,0.050887111160490245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,10240,0.04349866509437561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,10240,0.019932443896929424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,8192,0.009806222385830348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,7168,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,8192,0.017626666360431247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,8192,0.04056444433000352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,6144,0.010129777921570672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,7168,0.03605511122279697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,7168,0.016371554798550077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,5120,0.01013777818944719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,6144,0.01591822173860338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,6144,0.03460089034504361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,4096,0.007622222105662028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,5120,0.014871999621391296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,5120,0.03250755535231696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,4096,0.01386755539311303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,3584,0.0069982219073507525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,4096,0.029890666405359905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,3584,0.013558222187889947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,3072,0.006337777607970768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,3584,0.028845333390765723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,2560,0.005649777750174205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,3072,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,3072,0.02738044493728214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,2048,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,2560,0.02676711148685879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,2048,0.025452444950739544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,2048,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,1536,0.005903111149867375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,1536,0.024671999944580927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,1024,0.004606222112973531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,1536,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,1024,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,768,0.004265777766704559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,1024,0.023323555787404377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,768,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,512,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,2560,0.013094222380055321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,512,0.011407111254003314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,256,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,512,0.022340445054901972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,256,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,128,0.00315022220214208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,256,0.021659556362364028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,128,0.010762666662534079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,64,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,128,0.019711110326978896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,1024,32,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,64,0.010824888944625854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,65536,0.024297777149412367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,1024,768,0.022449778185950384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,1024,32,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,51200,0.020641777250501845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,65536,0.08102755414115058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,65536,0.1757084396150377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,16384,0.016456888781653512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,51200,0.0639102194044325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,51200,0.14279822508494058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,16384,0.05986933575736152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,16384,0.02682311170630985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,12288,0.014596444037225513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,12288,0.04982666505707634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,12288,0.02130488885773553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,10240,0.012279111478063794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,10240,0.043226665920681424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,10240,0.019311999281247456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,8192,0.010795555180973478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,8192,0.0387662218676673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,8192,0.017620444297790527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,7168,0.009690666364298927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,7168,0.03640444411171807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,7168,0.016568000117937725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,6144,0.008839999636014303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,6144,0.034325334760877825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,6144,0.015844444433848064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,5120,0.008352888955010308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,5120,0.03164444367090861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,5120,0.015200000670221118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,4096,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,4096,0.03017510970433553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,3584,0.006695111178689533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,4096,0.01443911095460256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,3584,0.013440888788965015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,3584,0.028568890359666612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,3072,0.006256888724035687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,3072,0.027834667099846735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,3072,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,2560,0.005551110953092575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,2560,0.02681155502796173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,2560,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,2048,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,2048,0.02548533346917894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,2048,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,1536,0.004650666481918759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,1536,0.024756444825066462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,1536,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,1024,0.003942222230964237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,1024,0.022350221872329712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,1024,0.012087999946541257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,768,0.003900444342030419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,768,0.02237066626548767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,512,0.003395555540919304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,768,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,512,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,512,0.02230311102337307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,256,0.0029306666304667792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,256,0.019638222124841478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,256,0.011519999967681037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,128,0.010760000182522668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,64,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,768,128,0.01947022146648831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,768,32,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,64,0.010757333702511258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,65536,0.018025777406162687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,768,32,0.01185066666867998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,65536,0.07826133569081624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,51200,0.015281778242852954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,65536,0.17631910906897652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,51200,0.06138222085105049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,16384,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,51200,0.14237244923909506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,16384,0.05957511398527357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,16384,0.02590755621592204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,12288,0.01072177787621816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,10240,0.009720000127951304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,12288,0.020612445142534044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,12288,0.04882133338186476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,8192,0.008041777544551426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,10240,0.042819556262758046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,8192,0.03858222232924567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,8192,0.017549332645204332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,7168,0.008037333687146505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,7168,0.03597955571280585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,6144,0.007666666474607255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,6144,0.015570667054918079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,6144,0.034239110019471906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,5120,0.008007110820876228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,10240,0.019319999549123976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,5120,0.014747555057207743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,4096,0.007337777978844113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,5120,0.03261955579121908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,4096,0.014003554979960123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,4096,0.030468444029490154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,3584,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,3584,0.028539554940329656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,3072,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,3072,0.013517333401574029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,7168,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,3072,0.02786488996611701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,2560,0.005519999812046687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,2560,0.02736888825893402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,2560,0.01314399970902337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,2048,0.005294222384691238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,2048,0.026112889250119526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,2048,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,1536,0.004559111264016894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,1536,0.012498666842778524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,3584,0.006623111251327727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,1024,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,1024,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,1024,0.022456000248591106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,768,0.0036319999231232535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,768,0.011485333244005838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,768,0.022443556123309668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,512,0.0032586666444937387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,512,0.011236444115638733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,512,0.021336000826623704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,256,0.003137777869900068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,256,0.010773333410422007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,256,0.019652444455358718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,128,0.002946666752298673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,1536,0.025416000021828547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,128,0.010712888505723743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,64,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,512,128,0.018953777021831937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,512,32,0.002642666714058982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,64,0.010762666662534079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,65536,0.017273777061038546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,512,32,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,65536,0.07511022355821398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,51200,0.014799111419253878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,65536,0.17479822370741102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,51200,0.060119999779595264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,51200,0.14189244641198054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,16384,0.009391999906963771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,16384,0.024877332978778418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,16384,0.05822844637764824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,12288,0.009656888743241629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,12288,0.020310221446885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,12288,0.04856266578038534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,10240,0.00867911097076204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,10240,0.019007111589113872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,8192,0.007956444389290279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,8192,0.01724088854259915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,8192,0.039084444443384804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,7168,0.007616889145639207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,7168,0.016207999653286405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,7168,0.03634577658441331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,6144,0.008030222521887885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,10240,0.042865776353412204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,6144,0.01551555593808492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,6144,0.03436711099412706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,5120,0.014513777361975776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,5120,0.03192355566554599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,4096,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,4096,0.013530666629473368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,4096,0.030187557140986126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,3584,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,5120,0.007655999726719326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,3584,0.013119999733236102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,3072,0.006592888798978593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,3584,0.02916533417171902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,3072,0.028156446086035833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,2560,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,2560,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,2560,0.027456889549891155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,2048,0.004928888960017098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,2048,0.012460444536474017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,2048,0.02602044410175747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,3072,0.013167999684810638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,1536,0.012072000238630505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,1536,0.02474311159716712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,1024,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,1024,0.011671110987663269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,1024,0.022437334060668945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,768,0.0038035557501845886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,1536,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,768,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,512,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,768,0.02334044377009074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,512,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,512,0.022336888644430373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,256,0.003118222165438864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,256,0.010371555884679159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,256,0.020620443754725985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,64,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,256,128,0.01756088932355245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,128,0.010416000253624385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,256,32,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,64,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,256,32,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,65536,0.07490222321616279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,51200,0.014448889427714877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,65536,0.17483199967278373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,51200,0.05863022141986423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,16384,0.009008888569143083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,16384,0.024831111232439678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,51200,0.14168532689412436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,12288,0.008012444608741337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,65536,0.01482222146458096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,12288,0.020668443706300523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,12288,0.04850666721661886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,10240,0.00869955536392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,10240,0.01868088874551985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,10240,0.04316622349951002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,8192,0.00795555528667238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,8192,0.017197334104114108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,16384,0.05820800198449028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,7168,0.007635555333561367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,7168,0.016238222519556682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,7168,0.03606399893760681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,6144,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,6144,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,6144,0.0347724457581838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,8192,0.03872799873352051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,5120,0.006656000183688269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,5120,0.032278223170174494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,4096,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,5120,0.014463999205165438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,4096,0.014098667436175875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,4096,0.030199110507965088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,3584,0.006262222097979651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,3584,0.029014223151736792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,3584,0.013482666677898832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,3072,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,3072,0.028488000233968098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,2560,0.006008000009589725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,3072,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,2560,0.012923555241690742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,2048,0.005994666781690385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,2560,0.026425777210129633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,2048,0.012452444268597497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,2048,0.025120000044504803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,1536,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,1536,0.024099555280473497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,1024,0.0036924444139003754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,1024,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,1536,0.004552000098758274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,768,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,1024,0.023433778021070693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,768,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,512,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,768,0.02277333372169071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,512,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,256,0.0029724445194005966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,512,0.021349334054523047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,256,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,256,0.020631111330456205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,128,0.010450666977299584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,80,128,128,0.018409777018758986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,64,0.0025902222841978073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,64,0.010804444551467896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,32,0.0025742221623659134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,128,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,65536,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,128,32,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,51200,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,65536,0.07403555843565199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,16384,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,51200,0.056820445590549044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,12288,0.007344889144102733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,12288,0.02072355482313368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,10240,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,10240,0.018971555762820773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,8192,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,16384,0.0251226673523585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,8192,0.017208000024159748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,7168,0.006344888773229387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,6144,0.006325333482689328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,7168,0.016185777054892648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,6144,0.015503999259736804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,5120,0.005960000058015187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,4096,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,5120,0.014848889576064216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,3584,0.005945777727497949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,4096,0.013793778088357715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,3072,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,3584,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,2560,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,3072,0.013167999684810638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,2560,0.013150222599506378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,2048,0.005659555395444234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,1536,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,2048,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,1024,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,1536,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,768,0.0032924444725116095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,1024,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,512,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,768,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,256,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,512,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,128,0.0027111110587914786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,256,0.010846222440401712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,128,0.01075644459989336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,64,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,32,0.002546666603949335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,64,32,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,65536,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,65536,0.07118222448560926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,51200,0.013180444637934366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,64,64,0.002551111082235972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,51200,0.05617244376076592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,16384,0.008001777860853408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,16384,0.025017778078715008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,12288,0.007400000260935889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,12288,0.020779555042584736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,10240,0.00700800038046307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,10240,0.019297776950730216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,8192,0.006689777804745569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,8192,0.01722755531469981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,7168,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,7168,0.01661333276165856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,6144,0.006299555715587404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,6144,0.01551555593808492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,5120,0.005954666684071223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,5120,0.014688000082969666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,4096,0.005631111148330901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,4096,0.014170666535695394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,3584,0.0063546668324205614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,3584,0.01349066694577535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,3072,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,3072,0.013463999662134381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,2560,0.006256000035338932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,2560,0.01310488912794325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,2048,0.005983110931184556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,1536,0.004384888956944148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,1536,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,1024,0.003719999972316954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,1024,0.011567999919255575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,768,0.003419555516706573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,768,0.011388444238238864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,512,0.0033484444850020935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,2048,0.012609778179062737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,512,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,256,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,128,0.00253866674999396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,256,0.010497777826256223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,128,0.010710222025712332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,64,0.010973333484596677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,32,0.002599999929467837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,80,32,32,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,80,32,64,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,51200,0.6132097774081761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,51200,1.1283022562662761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,16384,0.380021333694458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,16384,0.2036444478564792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,12288,0.2821715672810872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,12288,0.1566231118308173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,8192,0.17307021882798937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,10240,0.13413333892822266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,8192,0.1080720027287801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,7168,0.14723377757602268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,7168,0.09690311219957139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,6144,0.12986132833692762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,10240,0.21037511030832926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,5120,0.10684888892703587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,6144,0.08542933066685994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,4096,0.08768266439437866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,3584,0.07732622490988837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,4096,0.06137333313624064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,5120,0.07370755407545301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,3072,0.06746755705939399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,2560,0.05770044194327461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,3584,0.05554311143027412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,3072,0.050429334243138633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,2048,0.04698666599061754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,2560,0.044032888280020825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,1536,0.03604088889227973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,2048,0.038226667377683855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,1536,0.05176266696718004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,768,0.02050844497150845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,1024,0.03874666823281182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,512,0.014151111245155334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,768,0.03013422091801961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,256,0.009360888765917884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,512,0.023765333824687537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,128,0.0070088886552386815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,256,0.01866222255759769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,64,0.005062222066852781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,128,0.016889777448442247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,32,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,64,0.0203493336836497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,65536,32,0.0206631107462777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,65536,1024,0.025871111287011042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,65536,0.6277768876817491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,51200,0.8250684208340116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,65536,1.0487875408596463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,16384,0.2790542178683811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,16384,0.16386577818128797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,51200,0.48815732532077366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,12288,0.20696000258127847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,12288,0.1263715558581882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,10240,0.173689775996738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,8192,0.1436942285961575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,8192,0.08824444479412502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,7168,0.12068622642093235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,10240,0.10864888959460789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,6144,0.10841866996553208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,5120,0.08832444747289021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,7168,0.07939555247624715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,6144,0.07038755549324884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,4096,0.07342133257124159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,3584,0.06343111064698961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,4096,0.051627556482950844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,5120,0.06089244286219279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,3584,0.04692177640067207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,3072,0.05570222271813286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,2560,0.047172443734274976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,3072,0.04288711150487264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,2048,0.03894666830698649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,2560,0.037794666157828435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,1536,0.03022577696376377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,2048,0.033167110549079046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,1536,0.043656888935301036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,1024,0.021705778108702764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,768,0.016567111015319824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,1024,0.0318488876024882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,512,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,768,0.026449778013759192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,512,0.020957332518365648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,256,0.008358221915033128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,128,0.004939555707905027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,256,0.016718222035302054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,64,0.004601777841647466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,128,0.015928000211715698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,51200,32,0.004601777841647466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,64,0.018429333964983623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,51200,32,0.019281778070661757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,65536,0.32098134358723956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,65536,0.24485243691338432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,51200,0.24989867210388184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,16384,0.08915111091401841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,51200,0.19207999441358778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,12288,0.0643120010693868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,12288,0.055028445190853544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,10240,0.05465244584613376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,10240,0.04810488886303372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,8192,0.04560888806978861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,8192,0.04052977760632833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,16384,0.06896444161732991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,7168,0.039437333742777504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,6144,0.034867554903030396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,7168,0.03704266746838888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,5120,0.02986488739649455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,6144,0.03379377722740173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,4096,0.02422933280467987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,5120,0.030055999755859375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,3584,0.021718222233984206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,4096,0.02609777781698439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,3072,0.019279999865425956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,3584,0.02514933380815718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,2560,0.0166640000210868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,3072,0.022999111149046157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,2048,0.013651555611027611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,2048,0.018910222583346896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,1536,0.010976889067225985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,1536,0.021725333399242822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,1024,0.008450667063395182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,1024,0.019291554888089497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,768,0.0073902226156658595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,2560,0.020647111866209242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,512,0.005772444523043103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,768,0.017107556263605755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,256,0.0037502224246660867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,512,0.014517333772447375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,128,0.0035955554081334006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,256,0.012825777961148156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,64,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,128,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,16384,32,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,64,0.013117333253224691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,16384,32,0.012862222062216865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,65536,0.2568906678093804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,51200,0.19663466347588432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,16384,0.0745697749985589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,65536,0.19321955574883354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,12288,0.05387377738952637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,16384,0.05626844697528415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,10240,0.04652000135845608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,12288,0.04524088899294535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,8192,0.03888444436921014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,10240,0.039805332819620766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,7168,0.03324000040690104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,8192,0.03424266642994351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,6144,0.029529776838090684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,7168,0.031488888793521456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,5120,0.025424000289705064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,6144,0.02872266703181797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,4096,0.020948444803555805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,5120,0.025204444924990337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,3584,0.018719999326599967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,51200,0.15282311704423693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,4096,0.022218666142887537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,3584,0.021186666356192693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,3072,0.016346666547987197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,3072,0.019759999381171334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,2048,0.012323555847009024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,2560,0.01889511115021176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,1536,0.010424888796276517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,2048,0.017247110605239868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,1024,0.008147555920812819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,1536,0.020302222834693063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,768,0.007161777880456712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,1024,0.016555555992656283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,2560,0.014363555444611443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,512,0.005008888741334279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,768,0.015306666493415833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,512,0.013830222189426422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,256,0.003937777959638172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,128,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,256,0.01219733307758967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,64,0.0036159998012913596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,128,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,12288,32,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,64,0.0121760004096561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,65536,0.2207742267184787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,12288,32,0.012320000264379712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,65536,0.18692355685763887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,51200,0.14765244060092503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,16384,0.06312266985575359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,16384,0.05432000093989902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,12288,0.047373334566752114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,12288,0.04358844293488396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,51200,0.1695502201716105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,10240,0.039134220944510564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,8192,0.03212799959712558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,10240,0.03872088922394647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,8192,0.032829334338506065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,7168,0.02828177809715271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,6144,0.024878222081396315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,7168,0.029695110188590154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,5120,0.021547555923461914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,5120,0.024179556303554114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,4096,0.016731555263201397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,6144,0.02712000078625149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,3584,0.015232000086042615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,4096,0.021565333008766174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,3072,0.01424888935354021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,3072,0.019340443942281935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,3584,0.02051911089155409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,2560,0.01237422227859497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,2048,0.010827555424637265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,2560,0.018216000662909616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,2048,0.016945777667893302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,1536,0.008997333546479544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,1024,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,1536,0.018025777406162687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,768,0.006055111272467508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,512,0.00490044429898262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,1024,0.015841777125994366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,768,0.014504000544548035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,256,0.004055110944641961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,128,0.003937777959638172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,256,0.012049777640236748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,512,0.013270222478442721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,128,0.011808888779746162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,32,0.003630222131808599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,64,0.01146044416560067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,10240,32,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,65536,0.17094666428036162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,65536,0.14524266454908583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,51200,0.13346755504608154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,16384,0.05077066686418322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,51200,0.11493333180745442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,12288,0.037145776881111994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,10240,64,0.003549333247873518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,10240,0.031574222776624895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,12288,0.03619111246532864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,10240,0.0317848887708452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,8192,0.026583999395370483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,16384,0.043599999613232084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,8192,0.027075555589463975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,6144,0.020790222618314955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,7168,0.024933333198229473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,5120,0.018061333232455783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,6144,0.022520000735918682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,4096,0.014831999937693277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,5120,0.02092355489730835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,3584,0.01330933305952284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,4096,0.01903555625014835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,3072,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,3584,0.018473777506086562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,2560,0.01054666688044866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,3072,0.01763199932045407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,2048,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,7168,0.02314400010638767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,1536,0.008038221961922115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,2048,0.015406222807036506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,1024,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,2560,0.016591111818949383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,1536,0.016826666063732572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,768,0.0047084444926844705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,1024,0.01424533294306861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,512,0.004202666795916027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,256,0.0034737777378824023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,768,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,128,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,512,0.012774222426944308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,64,0.0031822222388452957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,128,0.011728000309732227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,8192,32,0.003012444410059187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,64,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,256,0.01180355581972334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,8192,32,0.012049777640236748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,65536,0.1573964489830865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,51200,0.11878132820129395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,65536,0.1393457783593072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,16384,0.045648889409171216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,51200,0.11129155423906113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,12288,0.03470755616823832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,16384,0.04258666767014397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,10240,0.030351112286249798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,12288,0.03465422325664096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,8192,0.025461332665549383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,10240,0.03117777903874715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,7168,0.02216177847650316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,8192,0.02609066665172577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,6144,0.01978488928741879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,7168,0.024111110303137038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,5120,0.01683911184469859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,6144,0.022227555513381958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,4096,0.014037332601017423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,5120,0.0206675562593672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,3584,0.012821333275900947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,4096,0.018927999668651156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,3072,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,3584,0.018204443984561496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,2560,0.010185778141021729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,3072,0.017543110582563613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,2048,0.009095999929640029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,2560,0.016221332881185744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,1536,0.007696000238259633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,2048,0.014466666513019137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,1024,0.005554666535721884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,1536,0.016872889465755887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,768,0.00453333349691497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,1024,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,512,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,512,0.012695111334323883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,256,0.0034640000926123727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,256,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,128,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,128,0.011337777806652917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,64,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,64,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,7168,32,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,32,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,7168,768,0.014021333720948962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,65536,0.1281920009189182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,51200,0.1008675562010871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,65536,0.13548621866438124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,16384,0.04020177655749851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,12288,0.03122666809293959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,16384,0.041723555988735624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,10240,0.0281368891398112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,12288,0.03402577837308248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,51200,0.10786933369106716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,8192,0.022104889154434204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,7168,0.019507555498017203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,10240,0.02956977817747328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,6144,0.01752800080511305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,7168,0.023443554838498432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,6144,0.021653332644038733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,5120,0.014757333530320061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,4096,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,8192,0.025991999440722998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,5120,0.02017866737312741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,4096,0.018872888551818002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,3584,0.011277333729796939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,3072,0.010059555371602377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,2560,0.009145778086450364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,3584,0.018147556318177115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,2048,0.008240888516108194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,2560,0.0155395550860299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,1536,0.00699733363257514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,2048,0.014130666851997375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,1024,0.00554666668176651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,3072,0.017743999759356182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,1536,0.016128000285890367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,768,0.004569777597983678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,1024,0.013762666947311826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,512,0.003938666648334927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,256,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,768,0.013576000101036496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,128,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,128,0.011712888876597086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,64,0.0031644445326593188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,256,0.011920889218648275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,64,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,6144,32,0.003335111050142182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,32,0.012216000093354119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,65536,0.1102364460627238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,51200,0.08929333421919082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,65536,0.13317155838012695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,16384,0.03493155704604255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,6144,512,0.012412444584899478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,12288,0.02737599942419264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,16384,0.040786667002571955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,51200,0.1059724423620436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,12288,0.03288088904486762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,8192,0.020640888147883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,8192,0.02459822263982561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,7168,0.018026666508780587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,10240,0.028547555208206177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,7168,0.023545778459972803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,6144,0.01578044394652049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,10240,0.024693333440356787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,5120,0.01367644468943278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,6144,0.0217902229891883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,4096,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,5120,0.020278222031063504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,3584,0.010403555300500657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,4096,0.01865511139233907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,3072,0.009756444229020013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,3584,0.0177875558535258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,2560,0.00868266655339135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,3072,0.016315556234783597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,2560,0.014863111906581454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,2048,0.007963555554548899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,1536,0.006587555425034628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,2048,0.014136888914638095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,1024,0.0052728888889153796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,1536,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,768,0.004872000051869286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,1024,0.013632000320487551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,512,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,768,0.013432000246312885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,256,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,512,0.012441777520709567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,128,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,256,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,64,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,5120,32,0.0032053333189752367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,64,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,5120,32,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,65536,0.09372622436947292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,51200,0.07111733489566378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,65536,0.09369688563876682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,16384,0.027986665566762287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,51200,0.0745697749985589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,16384,0.029655999607510034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,12288,0.02477333280775282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,10240,0.021547555923461914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,12288,0.024248000648286607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,8192,0.018212444252438016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,10240,0.02225422196918064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,7168,0.016926222377353244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,8192,0.019664888580640156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,6144,0.015261333849694995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,7168,0.01882488860024346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,6144,0.017907554904619854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,5120,0.014023999373118082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,4096,0.010523555179437002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,5120,0.01689511040846507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,3584,0.00980000032318963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,3584,0.014906666345066495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,4096,0.015855110353893705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,3072,0.008375110725561777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,2560,0.007436444362004598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,3072,0.013809777796268463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,2560,0.013457777599493662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,2048,0.006416000011894438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,1536,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,2048,0.013077333569526672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,1024,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,1536,0.013815999858909182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,768,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,1024,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,512,0.00360000009338061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,768,0.012845333251688214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,256,0.0033235556135574975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,512,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,128,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,256,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,128,0.011178666518794166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,64,0.0031840000301599503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,4096,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,64,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,65536,0.07787022325727674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,4096,32,0.01183555523554484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,51200,0.0624471108118693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,65536,0.09010133478376602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,16384,0.025576000412305195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,51200,0.07261955738067627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,12288,0.021839999490314062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,10240,0.019018666611777413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,10240,0.021553777986102637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,8192,0.016689777374267578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,12288,0.023748444186316595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,7168,0.015819556183285184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,8192,0.01940355532699161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,6144,0.014496000276671516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,7168,0.018598222070270114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,5120,0.01310488912794325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,6144,0.01756177842617035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,4096,0.01073066641887029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,5120,0.016344888342751395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,3584,0.009385777844323052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,4096,0.014984000060293408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,3072,0.008005333443482717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,3584,0.014183999763594734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,2560,0.006950221955776215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,3072,0.013773333695199756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,2560,0.013417777915795645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,2048,0.005952888892756567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,16384,0.02808711263868544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,2048,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,1536,0.004915555732117759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,1024,0.0039813332259655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,1536,0.013806222213639153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,1024,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,768,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,512,0.00360622215602133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,768,0.012647999657524956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,512,0.012149333126015134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,128,0.00327822214199437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,256,0.011220444407727985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,64,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,128,0.01146133326821857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3584,32,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,64,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,65536,0.06968000200059679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3584,32,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,51200,0.05534222059779697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,65536,0.08716711070802476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,16384,0.02293155590693156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,51200,0.0700133310423957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,16384,0.02681333323319753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,12288,0.017958222164048087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,10240,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,12288,0.0229102224111557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,8192,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,10240,0.02104355560408698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,7168,0.013154666456911298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,8192,0.01885777711868286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,6144,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,7168,0.01829688913292355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,5120,0.010418666733635796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,6144,0.017272889614105225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,4096,0.008791999684439765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,4096,0.014515555567211576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,5120,0.016179554992251925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,3584,0.008357333640257517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,3072,0.007730666961934831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,3584,0.014123555686738757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,3072,0.013811555173661975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,2560,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,2048,0.006335111127959357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,2048,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,1536,0.004935111022657818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,1536,0.013474666410022311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,1024,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,1024,0.012602667013804117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,2560,0.006800888727108638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,768,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,512,0.003645333151022593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,768,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,512,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,256,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,128,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,256,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,128,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,64,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,3072,32,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,65536,0.058871997727288135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,3072,32,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,51200,0.04761955473158094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,65536,0.08445600006315444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,16384,0.021054221524132624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,51200,0.06767555740144518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,16384,0.02640266716480255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,12288,0.022405332989162866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,10240,0.01516533394654592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,8192,0.013376889129479727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,10240,0.02055911057525211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,7168,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,8192,0.01867822143766615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,6144,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,7168,0.01770755483044518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,5120,0.00981155534585317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,6144,0.016906667086813185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,4096,0.008712000317043727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,5120,0.01551466683546702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,3584,0.008057778080304464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,12288,0.016422222057978313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,3072,0.0073315559162033936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,4096,0.014539556370841132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,2560,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,3584,0.014160000615649752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,2048,0.005656888915432825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,3072,0.01315822203954061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,1536,0.004639111045334074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,2560,0.013366222381591797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,2048,0.012815111213260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,1024,0.004222222086456087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,768,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,1536,0.013074666261672974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,512,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,1024,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,256,0.003185777821474605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,768,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,512,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,128,0.003024000053604444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,256,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,64,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,128,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2560,32,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,64,0.01147822207874722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2560,32,0.010440000229411654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,65536,0.04855377806557549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,51200,0.041857779026031494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,65536,0.08169777525795831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,16384,0.017095110482639737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,51200,0.06601333618164062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,16384,0.025498666697078284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,12288,0.014535999960369535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,10240,0.013368888861603208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,12288,0.022071111533376906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,8192,0.012240000069141388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,10240,0.0203164451652103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,7168,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,8192,0.01826844447188907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,6144,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,7168,0.01721600029203627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,6144,0.015512888630231222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,5120,0.009672889278994666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,4096,0.009733333355850643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,5120,0.014613333675596448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,3584,0.009684444301658207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,4096,0.014191110928853353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,3072,0.008999110923873054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,3584,0.013883555928866068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,2560,0.009294222626421187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,3072,0.013815110756291283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,2048,0.008912000391218398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,2048,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,1536,0.005728888842794631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,1536,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,1024,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,1024,0.012411555482281579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,768,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,2560,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,512,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,768,0.012130666938092975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,256,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,512,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,128,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,256,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,128,0.010800888968838586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,64,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,2048,32,0.003116444374124209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,64,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,65536,0.04054133428467645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,2048,32,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,51200,0.03276977936426798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,65536,0.07935200134913127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,16384,0.014883556299739413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,51200,0.06353511412938435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,12288,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,16384,0.025207999679777358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,10240,0.012822222378518848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,12288,0.021962665849261816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,8192,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,10240,0.019596444235907663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,7168,0.010864888628323873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,8192,0.01756444407833947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,7168,0.016499555773205228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,6144,0.010049777726332346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,5120,0.0094373333785269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,6144,0.01553244392077128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,4096,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,5120,0.014856000741322836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,3584,0.007650666766696506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,3584,0.013922666509946188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,3072,0.00666933341158761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,3072,0.01350933313369751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,2560,0.005995555470387141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,2560,0.013168888787428537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,2048,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,4096,0.013932444983058505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,1536,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,2048,0.012830221818553077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,1024,0.003858666867017746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,1536,0.012520000338554382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,768,0.003597333199448056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,1024,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,512,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,768,0.012110222544935016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,512,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,128,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,256,0.011150221857759686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,64,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,128,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,64,0.011526222030321756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1536,32,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,65536,0.029050668080647785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1536,32,0.01036977767944336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,51200,0.02438666588730282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,65536,0.0765377746687995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,51200,0.06003733476003011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,16384,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,16384,0.02481511069668664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,12288,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,12288,0.020671111014154222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,10240,0.010783111055692038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,10240,0.01866755551762051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,8192,0.009686222506894005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,8192,0.016572443975342643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,7168,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,7168,0.015929778416951496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,6144,0.010048888623714447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,6144,0.015266666809717814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,5120,0.010064889159467485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,5120,0.014655111564530266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,4096,0.008387555678685507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,4096,0.013850666582584381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,3584,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,3584,0.01386222243309021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,3072,0.006276444428496891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,3072,0.013131555583741931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,2560,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,2560,0.01316177762216992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,2048,0.006351111249791251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,2048,0.012520888613329994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,1536,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,1536,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,1024,0.005292444593376583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,1024,0.011786667009194693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,768,0.004270222038030624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,768,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,512,0.0038542221817705366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,512,0.011634666886594562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,256,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,256,0.01112711098459032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,128,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,128,0.010747555229398938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,64,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,64,0.010816000401973724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,1024,32,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,1024,32,0.01035288886891471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,65536,0.024950222836600408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,51200,0.020240889655219186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,65536,0.07422577672534518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,16384,0.01626666718059116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,51200,0.05826489130655924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,12288,0.013439999686347114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,16384,0.024113777610990737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,10240,0.013262222210566202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,12288,0.020256888535287645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,8192,0.010064000056849586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,7168,0.009655111365848118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,10240,0.01827822294500139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,6144,0.009688888986905416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,8192,0.01683111157682207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,5120,0.00831911133395301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,7168,0.015807110402319167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,6144,0.015139554937680563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,4096,0.007272889216740926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,5120,0.01475644442770216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,3584,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,4096,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,3584,0.013766222529941134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,2560,0.005568000177542369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,3072,0.013125333521101209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,2048,0.005193777796294955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,2560,0.013074666261672974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,1536,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,2048,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,1024,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,1536,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,768,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,1024,0.011765333513418833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,512,0.0034906665484110513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,768,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,256,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,512,0.011522666447692446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,128,0.0028275555620590844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,3072,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,256,0.011384888655609555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,64,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,128,0.010788444843557147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,768,32,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,64,0.011111111276679568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,768,32,0.010328888893127441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,51200,0.01759822169939677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,65536,0.07182577583524916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,16384,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,51200,0.05684177743064033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,12288,0.010404444403118556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,16384,0.023327999644809302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,10240,0.009083555804358589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,12288,0.01993955506218804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,8192,0.008349333372380998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,10240,0.018558222386572096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,7168,0.007672888537247975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,8192,0.016495111915800307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,6144,0.007541333635648091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,7168,0.015864888827006023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,5120,0.00795111142926746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,6144,0.015169777803950839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,65536,0.01738844480779436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,4096,0.007391110890441471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,5120,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,3584,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,3584,0.013818666338920593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,4096,0.013831111292044321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,3072,0.006432000133726332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,2560,0.005642666584915585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,2560,0.01290844463639789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,2048,0.005224888937340842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,3072,0.013248888982666863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,1536,0.004620444443490771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,2048,0.012501333322789935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,1024,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,1536,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,768,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,1024,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,512,0.0034844444857703317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,512,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,256,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,768,0.011838222543398539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,128,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,256,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,64,0.002777777819169892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,128,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,512,32,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,64,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,65536,0.017280000779363845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,512,32,0.01036533299419615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,65536,0.06912533442179362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,51200,0.0558168888092041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,16384,0.009752888646390703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,16384,0.022886221607526142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,12288,0.009727111293209923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,12288,0.019674667053752475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,10240,0.008660444782839881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,51200,0.014882667197121514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,10240,0.01791733337773217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,8192,0.008060444560315874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,7168,0.00871022211180793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,8192,0.016269332832760282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,6144,0.008368000388145447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,6144,0.014852444330851236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,7168,0.01573866605758667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,5120,0.008414222134484185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,4096,0.007299555672539606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,4096,0.013527111046844058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,5120,0.01424533294306861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,3584,0.006694222076071634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,3072,0.006232000059551663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,3072,0.01277511070171992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,3584,0.01348000019788742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,2560,0.00563822231358952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,2048,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,2560,0.0129120002190272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,2048,0.01218577805492613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,1536,0.004288889053795072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,1024,0.003999999827808804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,1536,0.012136889000733694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,768,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,1024,0.01166311071978675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,768,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,256,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,512,0.003440888805521859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,256,0.01076533314254549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,512,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,64,0.0026106666773557663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,256,32,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,128,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,64,0.010400888820489248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,256,32,0.010096888575288985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,51200,0.013532444834709167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,65536,0.06871377759509616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,51200,0.053559998671213783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,16384,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,12288,0.009047110875447592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,16384,0.02293066680431366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,10240,0.009006222089131674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,12288,0.020006222857369315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,10240,0.017885333961910672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,8192,0.01627466744846768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,7168,0.008714666797055138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,65536,0.014767999450365702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,6144,0.008354666332403818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,7168,0.015503999259736804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,5120,0.007640000018808577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,6144,0.015600888265503777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,4096,0.0069982219073507525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,5120,0.014223999447292753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,3584,0.00665155549844106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,4096,0.014208000567224292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,3072,0.00656355544924736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,3584,0.013501333693663279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,8192,0.007632888853549957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,2560,0.006258666515350342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,2048,0.004881777697139316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,3072,0.013491555220550962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,1536,0.004650666481918759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,2048,0.012159999873903064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,2560,0.012837332983811697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,1024,0.0036951113078329298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,1536,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,768,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,1024,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,768,0.01181333346499337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,512,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,256,0.0029120000286234748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,512,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,128,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,256,0.010785777535703448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,128,0.010449777874681683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,64,0.0025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,128,32,0.00253866674999396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,64,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,65536,0.01555288831392924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,128,32,0.010115555591053432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,51200,0.01424533294306861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,65536,0.06574489010704888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,16384,0.008725333544943068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,51200,0.05240622162818909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,16384,0.022755554980701868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,12288,0.007692444655630324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,10240,0.007301333049933116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,12288,0.019703111714786954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,8192,0.0074879998962084455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,10240,0.01826844447188907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,7168,0.007328888608349695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,8192,0.016223111086421542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,6144,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,7168,0.015896888242827523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,6144,0.014857777290874057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,5120,0.006671999891599019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,4096,0.006236444330877728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,5120,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,3584,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,4096,0.013544888959990608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,3072,0.005647999958859549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,3584,0.01353422221210268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,2560,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,2560,0.012757333616415659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,3072,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,2048,0.005683555371231503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,1536,0.004289777742491828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,1536,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,1024,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,2048,0.012479999827014076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,768,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,1024,0.011501332951916588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,512,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,768,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,256,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,512,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,128,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,256,0.010751999914646149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,128,0.010447999669445885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,64,0.002542222125662698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,64,32,0.002569777684079276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,64,0.010427555276287926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,65536,0.015560888581805758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,64,32,0.010125333236323463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,51200,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,65536,0.06411200099521212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,16384,0.008976000050703684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,16384,0.022727110319667395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,51200,0.05239555570814344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,12288,0.007638221813572778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,10240,0.007334222396214803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,12288,0.019951999187469482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,10240,0.01796444422668881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,8192,0.007655111451943715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,7168,0.007280000381999546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,8192,0.016226665841208566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,6144,0.007175111108356052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,5120,0.006728888799746831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,6144,0.01481066644191742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,4096,0.006205333189831839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,5120,0.014866666661368476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,4096,0.013415111435784234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,3584,0.0060159998635451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,3072,0.006062222023804982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,3584,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,2560,0.005954666684071223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,3072,0.01350044459104538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,2048,0.006075555665625467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,2560,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,1536,0.004273777620659934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,2048,0.012524444195959302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,1024,0.003930666794379552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,1536,0.01218933363755544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,7168,0.01571999986966451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,768,0.003408000080121888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,512,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,1024,0.011992000043392181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,256,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,512,0.011160888605647616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,768,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,128,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,64,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,256,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,128,0.010441777606805166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,65,32,32,0.0025066667132907445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,64,0.011232888533009423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,65,32,32,0.010471999645233154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,51200,0.5187990930345323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,51200,1.0710692935519748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,16384,0.36106488439771867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,16384,0.17458132902781168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,16384,0.2578115463256836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,12288,0.263570679558648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,12288,0.13534400198194715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,12288,0.20118755764431426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,51200,0.750059551662869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,10240,0.22063467237684461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,10240,0.11602044105529785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,8192,0.18209066655900744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,8192,0.09372978077994452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,10240,0.1725297768910726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,7168,0.1566151115629408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,7168,0.08387644423378839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,8192,0.14342933230929905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,6144,0.1398879951900906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,7168,0.1295928955078125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,6144,0.07424000236723158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,5120,0.11473777559068467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,5120,0.064246224032508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,6144,0.11531466907925075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,4096,0.09279821978674994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,4096,0.05439822210205925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,5120,0.10137066576215957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,3584,0.0809217757648892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,4096,0.0867591102917989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,3584,0.04929066697756449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,3072,0.07185955842336018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,3072,0.04456355505519443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,3584,0.08043466673956977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,2560,0.059471997949812144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,2560,0.03970311085383097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,3072,0.07406044668621488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,2048,0.04977600110901726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,2048,0.03454400102297465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,2560,0.065957334306505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,1536,0.03851644529236688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,2048,0.05871911181343926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,1536,0.05251377820968628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,1024,0.027790221903059218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,1024,0.023731556203630235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,768,0.02203111184967889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,1024,0.045928001403808594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,768,0.019689778486887615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,768,0.04087111022737291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,512,0.015250666273964776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,1536,0.02959200077586704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,256,0.00907022257645925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,512,0.017551110850440133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,512,0.0357582237985399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,128,0.006849777781301075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,256,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,256,0.032891554964913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,64,0.004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,128,0.013546666337384118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,65536,128,0.027230223019917805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,65536,32,0.00489688871635331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,64,0.015229332778188916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,65536,32,0.01851111153761546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,65536,0.5541057586669922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,65536,1.0488800472683377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,51200,0.8237039777967664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,51200,0.42444621192084414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,16384,0.27912534607781303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,16384,0.14334221680959067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,12288,0.20748533142937553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,12288,0.11154845025804307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,16384,0.22512978977627227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,10240,0.17274844646453857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,12288,0.17502310540941027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,10240,0.09580355882644653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,8192,0.14240088727739122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,8192,0.07818577686945598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,10240,0.15044177903069392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,7168,0.12007021903991699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,51200,0.648182233174642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,8192,0.12627111540900335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,7168,0.07052266597747803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,65536,0.8250746726989746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,6144,0.10564711358812119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,5120,0.08798844284481472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,6144,0.062322669559054904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,5120,0.054485334290398486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,4096,0.07284266418880887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,7168,0.11405777931213379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,6144,0.10291288958655463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,4096,0.046592000457975596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,5120,0.09006666474872166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,3584,0.06307111183802287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,3072,0.05488622188568115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,4096,0.07789599895477295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,3584,0.04256622327698601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,3072,0.038776000340779625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,2560,0.04662933283381992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,3584,0.07243021991517809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,3072,0.06560799810621473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,2560,0.034781333472993635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,2048,0.039341333839628435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,2560,0.059261335266960986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,1536,0.02991022335158454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,2048,0.03039111031426324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,2048,0.0540888872411516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,1024,0.021778666310840186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,1536,0.025974222355418738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,1024,0.02027555637889438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,768,0.016514667206340365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,1024,0.04136177897453308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,768,0.018037334084510803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,512,0.011982222398122152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,768,0.03705777724583944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,512,0.03368799885114034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,512,0.016507556041081745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,256,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,256,0.013214222258991666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,128,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,256,0.029500444730122883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,1536,0.04853777752982246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,64,0.004676444249020683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,128,0.01275644451379776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,51200,32,0.004927999857399198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,51200,128,0.02439911166826884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,64,0.014442667365074158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,51200,32,0.01572711103492313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,65536,0.341921779844496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,65536,0.20897421571943495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,51200,0.2623555660247803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,51200,0.1598844395743476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,16384,0.09011999766031902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,16384,0.05917155742645264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,65536,0.22889688279893664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,16384,0.07563644647598267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,12288,0.06441422303517659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,12288,0.047763556241989136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,51200,0.1802826722462972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,10240,0.05455466773774889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,12288,0.061609778139326304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,8192,0.04518311222394308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,10240,0.05607022179497612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,10240,0.041991998751958214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,8192,0.035419556829664446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,7168,0.0391884446144104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,7168,0.03259822063975864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,6144,0.034648888640933566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,8192,0.049535112248526685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,7168,0.04660177893108792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,6144,0.029860443539089624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,5120,0.02986577815479702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,6144,0.04355022311210632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,4096,0.024123556084103052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,5120,0.02683555583159129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,5120,0.04021688964631822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,3584,0.021677333447668288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,4096,0.03550311260753208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,4096,0.02368444369898902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,3072,0.019306666321224637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,3584,0.03463733196258545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,3584,0.02151288919978672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,2560,0.016414221790101793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,3072,0.020283554991086323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,2560,0.03128266665670607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,2560,0.018574222922325134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,2048,0.01346933344999949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,2048,0.02959377898110284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,2048,0.01773599949147966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,1536,0.016320889194806416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,3072,0.032587554719712995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,1024,0.0086666668454806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,1536,0.0278479986720615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,1024,0.013823111024167804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,768,0.007304888632562425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,768,0.012524444195959302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,1536,0.010611555642551847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,512,0.005953777581453323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,768,0.02439733346303304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,512,0.01237511138121287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,512,0.021703110800849065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,256,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,256,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,256,0.02134399943881565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,1024,0.02720622221628825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,128,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,64,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,16384,128,0.01832711034350925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,64,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,32,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,16384,32,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,65536,0.25434666209750706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,16384,128,0.00397866674595409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,65536,0.1812666654586792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,51200,0.1954195499420166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,51200,0.14143644438849554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,16384,0.07210044728385077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,16384,0.05259644322925144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,65536,0.19379733668433297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,51200,0.15436711576249865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,12288,0.05229599939452278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,16384,0.0670880013042026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,12288,0.04246222310596042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,10240,0.044679999351501465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,10240,0.037639111280441284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,12288,0.05520266625616285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,8192,0.03758488761054145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,10240,0.05032089021470812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,8192,0.03253600001335144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,7168,0.03237333231502109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,8192,0.04442666636572944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,7168,0.04229777720239428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,7168,0.030004445049497817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,6144,0.02872177627351549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,6144,0.0394720004664527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,5120,0.024704888463020325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,6144,0.02720533311367035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,5120,0.024123556084103052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,4096,0.020298666424221463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,5120,0.03670488794644674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,4096,0.02127911150455475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,3584,0.01825066738658481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,4096,0.033477332856920027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,3584,0.01997777819633484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,3072,0.015896000795894198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,3584,0.032925334241655134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,3072,0.01919911139541202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,2560,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,3072,0.03169066707293192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,2560,0.01813066667980618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,2048,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,2560,0.02998577886157566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,2048,0.016551999582184684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,1536,0.010035555395815108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,2048,0.028619554307725694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,1536,0.015542222393883599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,1536,0.02752977775202857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,1024,0.008127999802430471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,768,0.006673777682913675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,1024,0.012679999901188744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,1024,0.025399999486075506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,512,0.004464888738261329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,768,0.012508444488048553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,768,0.022348443667093914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,256,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,512,0.02168888847033183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,512,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,128,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,256,0.0206657780541314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,128,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,12288,128,0.018947554959191214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,64,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,12288,32,0.003522666792074839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,64,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,32,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,65536,0.21300444338056776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,12288,256,0.011360000405046674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,65536,0.17463466856214735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,51200,0.1679075558980306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,51200,0.13525956206851536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,16384,0.05759021970960829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,65536,0.1891120009952121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,16384,0.052319112751219005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,12288,0.044551998376846313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,51200,0.14466578430599636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,16384,0.06526577472686768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,10240,0.03582933213975694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,12288,0.043423109584384494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,12288,0.05304177933269077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,8192,0.029704888661702473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,10240,0.03714933329158359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,8192,0.03212711215019226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,10240,0.05053333441416422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,7168,0.027407111393080816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,8192,0.043191111750072904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,6144,0.024471999870406255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,7168,0.04135200050142076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,7168,0.02899822261598375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,5120,0.020072888996866014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,6144,0.02552088929547204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,6144,0.03788266579310099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,4096,0.016189333465364244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,5120,0.022945778237448797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,5120,0.03566933340496487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,3584,0.014254222313563028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,4096,0.020756444997257657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,4096,0.03280444608794318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,3584,0.019903111788961623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,3072,0.012650666965378655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,3584,0.03242311212751601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,2560,0.011007111105653973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,3072,0.01862044466866387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,3072,0.03149688906139798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,2048,0.009374221993817223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,2560,0.03046577837732103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,2048,0.028836445675955877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,2048,0.016696888539526198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,1536,0.008075555165608725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,1536,0.015025777949227227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,1536,0.027152000202072993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,1024,0.006879999819729064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,2560,0.017788444956143696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,1024,0.012595555848545499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,1024,0.025068443682458665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,768,0.005961777849329843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,768,0.012139555480745105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,512,0.003990222182538775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,768,0.022699554761250813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,512,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,256,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,512,0.021687999367713928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,256,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,256,0.020375111036830477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,128,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,10240,128,0.018279999494552612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,128,0.01109777804878023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,64,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,32,0.0035013332962989807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,10240,32,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,65536,0.17813689178890654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,10240,64,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,65536,0.1246088875664605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,51200,0.13639644781748453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,51200,0.09710044331020778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,65536,0.17280266020033094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,16384,0.04893155561553108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,16384,0.03900533252292209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,16384,0.06084533532460531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,51200,0.13929511441124812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,12288,0.033129778173234724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,12288,0.03221244282192654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,12288,0.05108177661895752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,10240,0.028770665327707928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,10240,0.028841776980294123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,8192,0.023618666662110224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,10240,0.0469359987311893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,8192,0.024712888730896845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,7168,0.02144355575243632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,8192,0.04159555501408047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,7168,0.022396443618668452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,6144,0.018935110833909776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,7168,0.039678222603268094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,6144,0.02051111062367757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,6144,0.03705866800414191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,5120,0.01628177695804172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,5120,0.034845332304636635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,5120,0.018922666708628338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,4096,0.013070222404268054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,4096,0.01779911087618934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,4096,0.031904889477623835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,3584,0.013377777404255338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,3584,0.01700799994998508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,3072,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,3584,0.03216177887386746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,3072,0.0162773331006368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,2560,0.010681777364677854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,3072,0.030823998981051978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,2560,0.015591111448076038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,2560,0.029147555430730183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,2048,0.009328888522254096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,2048,0.027280888623661462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,2048,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,1536,0.007615110940403408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,1536,0.02442755632930332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,1536,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,1024,0.005615111026499007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,1024,0.024933333198229473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,1024,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,768,0.0046471113132105935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,768,0.021996445126003687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,768,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,512,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,512,0.021363554729355708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,512,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,256,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,256,0.020285333196322124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,128,0.003387555480003357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,128,0.01073688848151101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,8192,128,0.01809777816136678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,256,0.0035377778112888336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,64,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,8192,32,0.0035546666218174826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,64,0.011015111373530494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,65536,0.1551013390223185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,65536,0.11366755432552761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,51200,0.12197600470648871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,51200,0.09054844246970283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,65536,0.17105510499742296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,16384,0.04692177640067207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,8192,32,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,16384,0.036665777365366616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,51200,0.138230217827691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,12288,0.03711377912097507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,16384,0.060221334298451744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,10240,0.02773422168360816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,12288,0.030469334787792627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,12288,0.050541334682040744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,8192,0.022732444935374792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,10240,0.04636089007059733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,10240,0.026975999275843304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,8192,0.02252977755334642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,8192,0.04200088977813721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,7168,0.02011022302839491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,7168,0.039437333742777504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,7168,0.022329777479171753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,6144,0.01796444422668881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,6144,0.03712888889842563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,5120,0.015507555670208402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,6144,0.020030222005314298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,5120,0.01830666595035129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,4096,0.012888000243239932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,4096,0.01720622181892395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,5120,0.035522666242387556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,3584,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,4096,0.03169599837727017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,3584,0.03173866536882188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,3072,0.010355555348926121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,3584,0.01702222228050232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,3072,0.01605866683853997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,2560,0.009091555244392818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,3072,0.03108533223470052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,2560,0.01516711049609714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,2048,0.00832444429397583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,2560,0.029697779152128432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,2048,0.013210666676362356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,2048,0.027069333526823256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,1536,0.0069724445541699724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,1024,0.005767999837795894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,1536,0.02439022229777442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,1536,0.012545777691735161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,1024,0.023010666171709698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,1024,0.012234666281276278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,768,0.004297777596447202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,512,0.003994666867785984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,768,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,768,0.023012444376945496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,256,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,512,0.021368889345063105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,512,0.011887110769748688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,256,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,128,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,256,0.0203128887547387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,64,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,128,0.011137777732478248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,7168,32,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,64,0.010737777584128909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,7168,128,0.019324445062213473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,7168,32,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,65536,0.12599021858639187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,51200,0.10067555639478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,65536,0.10532710949579875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,51200,0.08400089210934109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,16384,0.040688001447253756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,16384,0.034782220919926964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,65536,0.1694035530090332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,12288,0.03124533428086175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,51200,0.13711467054155138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,16384,0.05952711237801445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,12288,0.028287112712860107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,10240,0.02810488806830512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,12288,0.05049511128001743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,8192,0.0199955552816391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,10240,0.024832889437675476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,10240,0.04561511013242933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,7168,0.019924443629052904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,8192,0.02255288925435808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,7168,0.020311999652120803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,6144,0.017736888594097562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,7168,0.03975466556019253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,6144,0.01904977858066559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,5120,0.015285332997639975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,6144,0.03710577885309855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,5120,0.017824000782436795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,4096,0.012452444268597497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,5120,0.03498844305674235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,4096,0.016501333978441026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,3584,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,8192,0.04100355505943298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,3584,0.016021332807011075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,3072,0.009707555174827576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,3584,0.03118844495879279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,3072,0.015418666932317944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,3072,0.03014755580160353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,2560,0.008565333154466417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,2560,0.014123555686738757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,2560,0.028483556376563177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,2048,0.007651555869314406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,2048,0.012851555314328937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,2048,0.025840888420740765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,4096,0.0328800015979343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,1536,0.012400888734393649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,1024,0.0045422220395671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,1536,0.025736000802781846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,1024,0.012077333198653327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,768,0.004121777911980947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,768,0.011864888999197217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,768,0.02270222206910451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,512,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,1536,0.006835555450783835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,512,0.011355555719799466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,512,0.021345777644051447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,256,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,1024,0.023822221491071913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,256,0.011011555790901184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,256,0.021313778228229944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,128,0.010741333166758219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,6144,128,0.01939644416173299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,64,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,6144,32,0.002934222213096089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,32,0.01074222226937612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,6144,64,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,65536,0.10929422246085273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,65536,0.09896800253126357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,51200,0.08797244230906169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,65536,0.16834489504496256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,51200,0.07854311333762275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,16384,0.034824000464545354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,51200,0.13629955715603298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,16384,0.032469332218170166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,12288,0.02680177821053399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,16384,0.05898844533496433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,12288,0.04935466580920749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,12288,0.027064000566800434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,10240,0.02472266720400916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,10240,0.0455662210782369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,10240,0.023137778043746948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,8192,0.021201777789327834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,8192,0.041407111618253924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,7168,0.018274666534529794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,7168,0.01955200069480472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,8192,0.020649777518378366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,6144,0.0140684445699056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,7168,0.039157334301206805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,6144,0.018231110440360177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,5120,0.013321778012646569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,6144,0.036978665325376726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,5120,0.017159110969967313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,4096,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,5120,0.034840001000298396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,4096,0.016158221496476066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,3584,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,4096,0.03221244282192654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,3584,0.015239111251301236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,3584,0.03170311119821336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,3072,0.00923466682434082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,3072,0.03020177616013421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,3072,0.015455111861228943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,2560,0.008291555775536431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,2048,0.007298666569921706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,2560,0.02718844513098399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,2048,0.01258933295806249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,2560,0.012817777693271637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,1536,0.005564444594913059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,2048,0.02680088910791609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,1024,0.00425155543618732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,1536,0.02371911042266422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,1536,0.012141333685980903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,768,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,1024,0.023019555542204116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,1024,0.012168000141779581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,512,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,768,0.02197422252760993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,768,0.011737777955002256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,512,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,512,0.021320889393488567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,256,0.020258666740523446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,256,0.011509333219793109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,128,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,5120,128,0.017927999297777813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,128,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,64,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,32,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,64,0.011715555356608497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,5120,32,0.01074399964676963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,65536,0.09562133418189155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,65536,0.09006577730178833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,5120,256,0.003531555748648114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,51200,0.07027466429604425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,51200,0.07435733079910278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,65536,0.1680391099717882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,16384,0.027615111735132005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,51200,0.1357768906487359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,16384,0.03013955553372701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,12288,0.02218311031659444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,16384,0.05855377515157064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,10240,0.019001776973406475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,12288,0.049385776122411094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,12288,0.02410488824049632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,8192,0.015806222955385845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,10240,0.021838222940762837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,8192,0.01980888843536377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,7168,0.014135999812020196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,8192,0.04025422202216254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,7168,0.038579556677076556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,7168,0.018583999739752877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,6144,0.012888000243239932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,6144,0.017874666386180453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,6144,0.03607022100024753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,5120,0.01163111130396525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,10240,0.04514044523239136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,5120,0.01651822196112739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,4096,0.009751111268997192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,5120,0.03421599997414483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,4096,0.03131999903255039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,4096,0.015523556205961438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,3584,0.00904799997806549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,3584,0.030974222554100886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,3072,0.008603555460770925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,3584,0.013839999834696451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,3072,0.013733333183659447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,3072,0.03026755650838216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,2560,0.007721777591440413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,2560,0.02611911131276025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,2560,0.012838222086429596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,2048,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,2048,0.024975111087163288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,1536,0.004982222285535601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,1536,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,2048,0.00638488887084855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,1536,0.024150222539901733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,1024,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,1024,0.023034666975339253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,768,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,1024,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,768,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,512,0.0035982223020659555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,768,0.022337777747048274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,512,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,256,0.0033048888047536216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,256,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,256,0.02003466586271922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,128,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,128,0.010768888725174798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,64,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,128,0.019036443697081674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,64,0.010645333263609143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,4096,32,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,4096,512,0.021377777059872944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,65536,0.07690222395790948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,4096,32,0.010498666928874122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,51200,0.06271555688646105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,65536,0.08708266417185466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,51200,0.07184444533454047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,16384,0.025983111725913152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,65536,0.16741600301530626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,16384,0.029084444046020508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,51200,0.1358648935953776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,16384,0.0588959985309177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,12288,0.019723556107944913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,10240,0.017500445246696472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,12288,0.04916622241338094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,10240,0.021252445048756067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,8192,0.01514311134815216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,12288,0.02330488959948222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,10240,0.04505066739188301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,7168,0.01310755560795466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,8192,0.018943111101786297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,8192,0.04016622238688999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,6144,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,7168,0.018333334061834548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,7168,0.03844711184501648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,5120,0.011151110960377587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,6144,0.01688088807794783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,6144,0.0362062222427792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,4096,0.009472000102202097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,5120,0.016197333733240765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,5120,0.03394044438997904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,3584,0.00830577810605367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,4096,0.014808000789748298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,4096,0.030886222918828327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,3072,0.007688889073001013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,3584,0.029564443561765883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,3584,0.014392889208263822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,3072,0.013756444056828817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,2560,0.007622222105662028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,3072,0.028528889020284016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,2560,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,2560,0.026072000463803608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,2048,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,2048,0.024700444605615404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,1536,0.004907555464241239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,1536,0.023750222391552393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,1536,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,1024,0.0041235557032956015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,1024,0.02271466619438595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,1024,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,768,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,768,0.011537777880827585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,512,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,2048,0.005650666852792104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,512,0.021983999345037673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,256,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,512,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,256,0.011156444748242697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,256,0.020054222808943856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,128,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,128,0.018990221950742934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3584,768,0.021949332621362474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,64,0.00313688897424274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,64,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3584,32,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,32,0.010403555300500657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,65536,0.0691937804222107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3584,128,0.010694222317801582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,65536,0.08652355273564656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,51200,0.05524444580078125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,51200,0.06968177689446343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,65536,0.1668186717563205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,16384,0.025136889682875738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,16384,0.027573333846198186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,51200,0.13596799638536242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,12288,0.01903022163444095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,12288,0.022630222969584998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,16384,0.05794577466117012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,10240,0.01752000053723653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,12288,0.049160000350740224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,10240,0.02091466718249851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,8192,0.015118221441904703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,10240,0.04469155602984958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,8192,0.03996888796488444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,8192,0.018608000543382432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,7168,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,7168,0.03827822208404541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,7168,0.018024888303544786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,6144,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,6144,0.03592533204290602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,6144,0.016712889075279236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,5120,0.010223111344708337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,5120,0.03393066591686673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,5120,0.015568888849682279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,4096,0.008718222379684448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,4096,0.03048711021741231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,4096,0.013785777820481194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,3584,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,3584,0.02983022067281935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,3072,0.00737955586777793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,3072,0.013531555732091268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,3584,0.008168000313970778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,3072,0.027914666467242773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,2560,0.006956444846259222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,2560,0.02570844524436527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,2560,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,2048,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,2048,0.024449777272012498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,2048,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,1536,0.004624888714816836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,1536,0.023789332972632513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,1024,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,1536,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,1024,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,1024,0.023062222533755835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,768,0.003937777959638172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,768,0.02197244432237413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,768,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,512,0.003814222084151374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,256,0.003522666792074839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,512,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,512,0.021651556094487507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,128,0.003124444435040156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,256,0.0199635558658176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,256,0.011367999845080905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,128,0.010751999914646149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,64,0.002959111084540685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,3072,32,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,64,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,3072,32,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,65536,0.05760800176196628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,65536,0.0842648877037896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,51200,0.04704088966051737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,65536,0.1670515537261963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,3072,128,0.017866666118303936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,51200,0.06786933210160997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,16384,0.022631110416518316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,51200,0.13562844859229192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,12288,0.017862222260899015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,16384,0.05780533287260267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,16384,0.026242666774325903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,10240,0.016336888074874878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,12288,0.022364444202846948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,12288,0.04859111044141981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,8192,0.013775111072593264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,10240,0.020376889242066275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,10240,0.04458311200141907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,8192,0.018633777896563213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,7168,0.013454222016864352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,8192,0.040383110443751015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,6144,0.012083555261294046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,7168,0.01760444376203749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,7168,0.038141333394580416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,5120,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,6144,0.034338666333092585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,6144,0.01690933273898231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,4096,0.008739555875460306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,5120,0.015184889237085978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,4096,0.014031999640994601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,3584,0.007832888927724626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,4096,0.02920711040496826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,3584,0.0138586668504609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,3584,0.029528889391157363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,3072,0.006696888970004187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,3072,0.01350933313369751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,5120,0.03362933463520474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,2560,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,3072,0.028548445966508653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,2560,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,2048,0.005261333452330695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,2560,0.026521777113278706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,2048,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,1536,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,2048,0.02519911030928294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,1536,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,1536,0.024375110864639282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,1024,0.004328888737493091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,1024,0.02274311085542043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,768,0.0038497779104444715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,1024,0.01201600001917945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,768,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,768,0.023027555810080633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,512,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,512,0.021295999487241108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,256,0.003301333429084884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,512,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,256,0.010869333313571082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,256,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,128,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2560,128,0.017871111631393433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,128,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,64,0.0028275555620590844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2560,32,0.0031822222388452957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,64,0.01148799972401725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2560,32,0.010449777874681683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,65536,0.047613332668940224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,65536,0.08107733064227633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,51200,0.04003377755482992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,65536,0.16775733894772002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,51200,0.06533777713775635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,16384,0.015440889530711703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,51200,0.13579022884368896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,16384,0.025782222549120586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,12288,0.014845333165592618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,16384,0.0586951110098097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,12288,0.021987555755509272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,10240,0.013397333522637686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,12288,0.04929866724544101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,10240,0.02000266644689772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,8192,0.011996444728639392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,10240,0.04495022363132901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,8192,0.017839110559887357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,8192,0.04028977619277106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,7168,0.011019555230935415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,6144,0.010415111151006486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,7168,0.03671288821432326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,7168,0.017488888568348356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,6144,0.015526221858130561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,5120,0.009649777577983009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,6144,0.03624711102909512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,5120,0.01510933372709486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,4096,0.008595555192894405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,5120,0.032010667853885226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,3584,0.007407111426194509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,4096,0.013807111316257052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,4096,0.03019022279315525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,3072,0.006676444576846228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,3584,0.013533333109484779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,3584,0.02875466810332404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,2560,0.005917333480384614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,3072,0.013407111167907715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,3072,0.027215111586782668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,2560,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,2048,0.005392888767851724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,2560,0.026483555634816486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,1536,0.004631999880075455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,2048,0.012447111308574677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,2048,0.025439111722840205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,1024,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,1536,0.023777777949968975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,1536,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,1024,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,768,0.003927999900446998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,1024,0.022859555151727464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,512,0.0032826666202810076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,768,0.0229751103454166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,512,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,256,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,512,0.02233422299226125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,256,0.010796444283591377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,256,0.021346666746669348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,128,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,2048,128,0.01795022189617157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,64,0.0028044444819291434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,128,0.01144177797767851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,2048,32,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,64,0.010381333529949188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,65536,0.040215111441082425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,32,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,2048,768,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,51200,0.033321777979532875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,65536,0.07709066735373603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,65536,0.1677208873960707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,51200,0.06344355477227105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,16384,0.01516533394654592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,51200,0.1365786658393012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,16384,0.02537600033813053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,12288,0.012585777375433179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,16384,0.05868444177839491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,12288,0.02167644434505039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,12288,0.04892533355289035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,10240,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,10240,0.043453332450654775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,10240,0.01993600030740102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,8192,0.010658666491508484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,8192,0.0377439988984002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,8192,0.01791911158296797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,7168,0.010016000105275048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,7168,0.016569778323173523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,7168,0.037008888191647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,6144,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,6144,0.015120888749758402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,5120,0.009092444347010719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,6144,0.03534222311443753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,5120,0.014490667316648694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,5120,0.03232622146606445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,4096,0.008001777860853408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,4096,0.028679112593332928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,4096,0.013970666461520724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,3584,0.007258666886223688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,3584,0.028447998894585505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,3072,0.006421333385838403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,3072,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,3072,0.027492443720499676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,2560,0.005922666854328579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,2560,0.01294844432009591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,2560,0.025832888152864244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,2048,0.005320000151793162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,3584,0.013536888692114087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,2048,0.012489777472284106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,1536,0.004517333375083076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,2048,0.025391111771265667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,1536,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,1024,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,1536,0.024346666203604803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,1024,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,1024,0.023795555035273235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,768,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,768,0.02167466613981459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,768,0.011807999677128263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,512,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,512,0.021008888880411785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,512,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,256,0.003265777809752358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,256,0.019976889093716938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,256,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,128,0.002809777855873108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1536,128,0.017873777283562552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,64,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,128,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1536,32,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,64,0.010436444646782346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,65536,0.026692445079485577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1536,32,0.011354666617181567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,65536,0.0749617748790317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,51200,0.02269777821169959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,65536,0.16848799917432997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,51200,0.05971999963124593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,16384,0.011407111254003314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,51200,0.13599110974205866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,16384,0.024880889389250014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,12288,0.01035733355416192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,16384,0.058556444115108915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,12288,0.020636444290479023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,12288,0.04898133211665683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,10240,0.010323555933104621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,10240,0.04219288958443535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,10240,0.01930488811598884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,8192,0.009016888837019602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,8192,0.03870577613512675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,7168,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,7168,0.01590044465329912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,8192,0.017476444443066914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,6144,0.007713778151406183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,7168,0.036010666026009455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,6144,0.015533333023389181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,6144,0.03355733222431607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,5120,0.007905777957704332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,5120,0.014473777678277759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,5120,0.03234044379658169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,4096,0.007335999773608313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,4096,0.028460443019866943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,4096,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,3584,0.006974221931563483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,3584,0.013510222236315409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,3584,0.02852177619934082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,3072,0.006575999988449945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,3072,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,2560,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,3072,0.02813777658674452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,2560,0.012792000340090858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,2560,0.026439110438028973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,2048,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,2048,0.01273333364062839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,2048,0.02515022291077508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,1536,0.004558222161398994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,1536,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,1536,0.024711110525661047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,1024,0.004053333153327306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,1024,0.012039999994966718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,1024,0.023919999599456787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,768,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,768,0.01181688904762268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,768,0.02269422180122799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,512,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,512,0.011716444459226398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,512,0.022338666849666174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,256,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,256,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,256,0.021338666478792827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,128,0.0028124445428450904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,128,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,1024,128,0.019294222195943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,64,0.002961777771512667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,64,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,1024,32,0.00267999991774559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,1024,32,0.011365333365069495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,65536,0.025231111380789015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,65536,0.07422400183147855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,65536,0.16779822773403594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,51200,0.021173333128293354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,51200,0.05882044633229574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,16384,0.010839111275143094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,51200,0.13527022467719182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,16384,0.0240666667620341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,12288,0.009511999785900116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,16384,0.05700622002283732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,10240,0.00981244444847107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,12288,0.02020711037847731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,12288,0.04721777637799581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,8192,0.008382221890820397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,10240,0.018466666340827942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,10240,0.04385777645640903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,7168,0.00795022232664956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,8192,0.0381626652346717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,8192,0.016536888149049547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,7168,0.016297777493794758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,6144,0.007883555359310573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,7168,0.035776889986462064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,5120,0.008070222205585903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,6144,0.015484443969196744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,6144,0.033372445238961115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,4096,0.0076657773719893555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,5120,0.014583110809326172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,5120,0.0324248903327518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,3584,0.0070293330483966405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,4096,0.029489778810077246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,3584,0.013576000101036496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,4096,0.013632888595263163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,3072,0.00673333348499404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,3584,0.028168890211317275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,2560,0.006552000012662675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,3072,0.028276443481445312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,3072,0.012830221818553077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,2560,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,2048,0.0053075556125905775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,2560,0.026456000076399908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,1536,0.004555555681387584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,2048,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,2048,0.024475556280877855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,1024,0.003863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,1536,0.0251226673523585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,1536,0.01240711079703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,768,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,1024,0.0236835562520557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,1024,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,512,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,768,0.02302933401531643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,512,0.01105955574247572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,512,0.021347555849287245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,256,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,256,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,256,0.020311110549502902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,128,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,768,128,0.018252443936136033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,768,0.011923555698659686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,64,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,32,0.0025928889711697898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,64,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,65536,0.018235555953449674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,768,32,0.010448888772063784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,768,128,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,65536,0.07088000244564481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,51200,0.01517511076397366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,65536,0.16668532954321968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,16384,0.010402667025725046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,51200,0.05755199988683065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,51200,0.13478667206234402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,12288,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,16384,0.023249778482649062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,16384,0.05700266361236572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,12288,0.0199626667631997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,10240,0.009069333473841349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,12288,0.04621866676542494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,10240,0.018592000007629395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,8192,0.008012444608741337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,10240,0.04220444295141432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,8192,0.016859556237856548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,8192,0.037420445018344455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,7168,0.007733333441946242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,7168,0.03564000129699707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,7168,0.015571556157535978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,6144,0.007274666594134436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,6144,0.03362400001949734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,6144,0.01609244445959727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,5120,0.007854222423500484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,5120,0.03161244591077169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,5120,0.014175999495718213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,4096,0.0070364442136552595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,4096,0.02983377708329095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,3584,0.0069253332912921906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,3584,0.013426666458447775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,3584,0.028360890017615423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,3072,0.006623111251327727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,3072,0.013175110850069257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,3072,0.02753422160943349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,2560,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,4096,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,2560,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,2048,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,2560,0.025740444660186768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,2048,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,1536,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,1536,0.012039110892348819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,1536,0.02511466708448198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,1024,0.0038559999730851916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,1024,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,1024,0.023731556203630235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,768,0.00350577798154619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,2048,0.024694222542974684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,768,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,512,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,768,0.021734222769737244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,512,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,256,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,512,0.021317332983016968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,256,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,128,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,256,0.01997866729895274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,128,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,512,128,0.018255111243989732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,64,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,512,32,0.0029502221279674103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,65536,0.014694222145610385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,32,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,512,64,0.010344888601038191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,65536,0.06920355558395386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,51200,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,65536,0.16651644971635607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,16384,0.00943377779589759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,51200,0.055925336149003774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,16384,0.023342221975326538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,51200,0.13462400436401367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,16384,0.05544355511665344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,12288,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,10240,0.007643555601437886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,12288,0.046019554138183594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,10240,0.01854311095343696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,12288,0.01998666591114468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,10240,0.04185333185725742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,8192,0.016548444827397663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,8192,0.03738222188419766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,7168,0.006934222247865465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,7168,0.015280889140235053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,7168,0.035678221119774714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,6144,0.0064239998658498125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,6144,0.033329778247409396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,6144,0.014985778265529208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,5120,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,5120,0.031296888987223305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,5120,0.014189332723617554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,4096,0.006849777781301075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,8192,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,3584,0.006581333362393909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,4096,0.01349066694577535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,3584,0.013396444420019785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,3072,0.006254222244024277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,3584,0.028464890188641016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,3072,0.013078221844302284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,2560,0.0062871111763848206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,3072,0.027456000447273254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,2560,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,2560,0.025748444928063288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,2048,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,2048,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,2048,0.025751110580232408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,1536,0.004416000097990036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,1536,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,4096,0.02920177910063002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,1024,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,1024,0.011687111523416309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,1024,0.023745778534147475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,1536,0.023924445112546284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,768,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,768,0.023031110564867657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,512,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,768,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,512,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,256,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,256,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,256,0.020463999774720933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,128,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,128,0.011139555937714048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,128,0.018938667244381376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,64,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,64,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,256,32,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,256,512,0.021663111117151048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,65536,0.014183999763594734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,256,32,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,65536,0.1664773358239068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,51200,0.012311999996503195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,65536,0.06877155436409844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,51200,0.0551111102104187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,16384,0.007668444679843054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,51200,0.13380889097849527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,16384,0.02327999969323476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,12288,0.0075235557225015424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,16384,0.05544355511665344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,12288,0.020023110840055678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,10240,0.0070675553547011475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,12288,0.047671112749311656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,10240,0.018631110588709515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,8192,0.007171555525726742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,10240,0.04209511147605049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,7168,0.00630044440428416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,8192,0.016711110870043438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,8192,0.03769333495034112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,6144,0.006633777585294511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,7168,0.035707556539111666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,7168,0.015871110889646742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,5120,0.006007110906971826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,6144,0.03398933344417148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,6144,0.015501333607567681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,4096,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,5120,0.014391111003028022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,5120,0.031672000885009766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,3584,0.006281777802440856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,4096,0.013810666898886362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,4096,0.02885066800647312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,3584,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,3072,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,3584,0.029527998632854883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,3072,0.01309511148267322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,2560,0.006335999816656113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,3072,0.027787556250890095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,2048,0.005680888891220093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,2560,0.012917333179050021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,2560,0.02644888891114129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,1536,0.004576000074545543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,2048,0.012356444365448423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,2048,0.025776889589097764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,1024,0.003970666478077571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,1536,0.01239999963177575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,1536,0.023733332753181458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,1024,0.012096888489193387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,768,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,1024,0.0240666667620341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,768,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,512,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,768,0.02240177823437585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,256,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,512,0.011425777441925473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,512,0.022408889399634466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,128,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,256,0.010808889236715106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,128,0.011374221907721626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,128,0.01790844400723775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,64,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,128,32,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,64,0.010472888747851053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,128,32,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,65536,0.010677333507272931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,51200,0.009666666388511658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,65536,0.06648177570766874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,51200,0.053095112244288124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,16384,0.00701422244310379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,12288,0.006633777585294511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,64,128,256,0.020644444558355544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,10240,0.006121777825885349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,16384,0.023026666707462732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,12288,0.020243555307388306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,8192,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,10240,0.018574222922325134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,7168,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,6144,0.00591466658645206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,8192,0.016879111528396606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,7168,0.015855110353893705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,5120,0.006304889089531369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,4096,0.006242666807439592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,6144,0.01483022173245748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,5120,0.014512888259357877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,3584,0.006265777680608962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,4096,0.014070222775141398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,3584,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,3072,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,2560,0.006232888748248418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,2560,0.01293155550956726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,2048,0.005640888793600931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,2048,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,1536,0.004224888980388641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,1536,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,3072,0.00598933340774642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,1024,0.0036435553597079385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,768,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,768,0.011788444386588203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,512,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,1024,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,256,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,512,0.011711999773979187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,128,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,128,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,256,0.011487111449241638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,64,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,64,32,0.002566222308410539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,32,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,65536,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,64,64,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,51200,0.009710222482681274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,65536,0.06627021895514594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,16384,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,51200,0.052743110391828746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,16384,0.023119111855824787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,12288,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,10240,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,12288,0.020062221421135795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,8192,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,10240,0.018516444497638278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,7168,0.006610666712125142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,8192,0.01682133310370975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,6144,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,7168,0.016180444094869826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,6144,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,5120,0.006248888870080312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,4096,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,5120,0.014544000228246054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,3584,0.006085333310895496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,3584,0.013473778135246701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,4096,0.01425333321094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,3072,0.00599911105301645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,2560,0.005985777825117111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,3072,0.013418667018413544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,2048,0.005897777775923411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,2048,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,1536,0.004259555704063839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,1536,0.012461333639091916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,1024,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,1024,0.012161778079138862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,2560,0.012764444781674279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,768,0.003731555408901639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,512,0.0033768889390759995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,768,0.011716444459226398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,256,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,256,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,128,0.0028435554769304064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,512,0.011818666425016193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,64,0.00300799993177255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,128,0.011062222222487131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,64,32,32,0.0027182222240500977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,64,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,64,32,32,0.010205333431561789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,16384,0.36282043986850315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,51200,0.5126621988084581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,51200,1.0680276023017037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,16384,0.17248000038994682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,10240,0.22104444768693712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,12288,0.2639848921034071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,12288,0.13362044758266872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,8192,0.18076444996727836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,10240,0.11451021830240886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,8192,0.0926462213198344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,7168,0.15550933943854436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,7168,0.08315733406278822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,6144,0.13663466771443686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,5120,0.11323289076487224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,4096,0.09047377771801418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,5120,0.06400977902942233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,3584,0.08081333504782783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,6144,0.0739884442753262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,3584,0.04929511083496941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,4096,0.053929779264662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,3072,0.06991022162967257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,2560,0.06031644344329834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,2048,0.04955910974078708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,3072,0.044106665584776134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,1536,0.03822844558291965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,2048,0.03466044531928168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,1024,0.028135998381508723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,1536,0.02981511089536879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,768,0.02159022291501363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,1024,0.02348622183005015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,768,0.02054222259256575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,512,0.01517777807182736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,512,0.017456000049908955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,256,0.00922400007645289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,256,0.015217777755525378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,128,0.0063662222690052455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,2560,0.03990222348107232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,128,0.014139556222491793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,64,0.0074746666683091064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,65536,32,0.007376888559924231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,64,0.014504889647165934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,65536,32,0.01868977811601427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,65536,1.0474347008599174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,65536,0.5459698041280111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,51200,0.8203777737087674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,51200,0.41849154896206325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,16384,0.27505866686503094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,16384,0.1419315603044298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,12288,0.20425866709815133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,10240,0.17079199684990776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,12288,0.11055022478103638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,8192,0.14223733213212755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,10240,0.09511555565728082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,7168,0.1196933322482639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,8192,0.0773493316438463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,7168,0.06979288657506307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,6144,0.1059066653251648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,6144,0.06182399723264906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,4096,0.07154577970504761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,5120,0.0541013346778022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,4096,0.04635733366012573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,3584,0.06349955664740668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,3072,0.05466933382882012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,3584,0.04257511099179586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,2560,0.046479112572140165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,2560,0.03460089034504361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,3072,0.0384133325682746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,2048,0.039621333281199135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,1536,0.02993333339691162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,5120,0.08814311027526855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,1024,0.021595555875036452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,2048,0.030352890491485596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,1536,0.026224889689021643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,768,0.016662221815851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,512,0.012056888805495368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,1024,0.01981244484583537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,768,0.019107555349667866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,256,0.008319999608728621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,128,0.004962666581074397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,512,0.016564443707466125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,256,0.014112888111008538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,64,0.00453244439429707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,51200,32,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,128,0.012863111164834766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,64,0.013796444568369122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,51200,32,0.017047999633683097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,65536,0.34058666229248047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,65536,0.19844888316260445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,16384,0.08841777510113186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,51200,0.26025154855516225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,12288,0.06345511145061918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,51200,0.15426399972703722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,16384,0.057814220587412514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,10240,0.0536026656627655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,8192,0.044285333818859525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,10240,0.04108622339036729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,12288,0.046725332736968994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,7168,0.03883822096718682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,6144,0.03436622354719374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,8192,0.034974220726225115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,7168,0.03229066729545593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,5120,0.029437333345413208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,6144,0.02955199943648444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,4096,0.023796444137891132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,3584,0.02161955502298143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,4096,0.023911111884646948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,5120,0.0274000002278222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,3072,0.01891644464598762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,2560,0.01626844373014238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,3584,0.022039999564488728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,2048,0.013519110778967539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,3072,0.02020355562369029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,2560,0.019112888309690688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,1536,0.010837333069907295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,1024,0.008567111359702216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,1536,0.01676977839734819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,2048,0.017596445149845548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,768,0.0074684446056683855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,512,0.005493333356248007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,1024,0.014392889208263822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,256,0.003843555433882607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,768,0.013559111290507846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,512,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,128,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,64,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,256,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,16384,32,0.0032791110376516977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,128,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,64,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,16384,32,0.013136000268989138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,65536,0.2542071077558729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,51200,0.1989306608835856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,65536,0.17289866341484916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,16384,0.07514222462972005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,51200,0.13536889023251003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,16384,0.051238222254647144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,12288,0.05204000075658163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,10240,0.04463466670778063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,12288,0.041489776637819074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,8192,0.03792000148031447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,8192,0.03198400139808655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,7168,0.032058666149775185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,7168,0.02944444285498725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,6144,0.028593778610229492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,6144,0.02695111102528042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,5120,0.02459999918937683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,10240,0.0370053317811754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,4096,0.01998133295112186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,5120,0.023864888482623633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,4096,0.021337777376174927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,3584,0.017997332745128207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,3072,0.015951999359660678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,3584,0.020250666472646926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,2560,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,3072,0.018686221705542672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,2048,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,2560,0.018256000346607633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,1536,0.009910222556855943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,2048,0.016923555069499545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,1024,0.008015111088752747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,1536,0.015807110402319167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,768,0.006722666737106111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,1024,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,768,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,512,0.0042275554604000514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,256,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,512,0.012138666378127204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,128,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,256,0.011619555453459421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,64,0.0033084443873829315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,128,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,12288,32,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,64,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,12288,32,0.012816000315878125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,65536,0.21171466509501138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,51200,0.16723644733428955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,65536,0.16803377204471162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,16384,0.057032889790005155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,16384,0.049772444698545665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,12288,0.04105955693456862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,51200,0.13056088818444148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,10240,0.03501510951254103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,8192,0.029367112451129492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,12288,0.040510220660103693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,7168,0.025677333275477093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,10240,0.03636711173587375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,6144,0.023017777336968318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,8192,0.030953778160942927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,5120,0.019679110911157396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,7168,0.02854666776127285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,4096,0.015977778368526034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,6144,0.025077333052953083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,5120,0.022672888305452134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,3584,0.014470222923490735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,3072,0.012850667039553324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,4096,0.02087200019094679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,2560,0.011349333657158745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,3584,0.01960533360640208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,2048,0.009693333672152625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,3072,0.01852799952030182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,1536,0.00831200016869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,2048,0.016183111402723525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,1024,0.006573333508438534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,1536,0.014137778017255994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,768,0.004691555682155821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,1024,0.012764444781674279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,512,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,768,0.012858666479587555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,256,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,512,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,128,0.003265777809752358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,256,0.011376000112957425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,64,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,128,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,10240,32,0.00349866681628757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,64,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,32,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,10240,2560,0.017982222967677645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,65536,0.17757156160142687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,51200,0.1356924507353041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,65536,0.11785777409871419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,16384,0.04846044381459554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,12288,0.03655822078386942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,51200,0.09206400314966838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,16384,0.03766399953100417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,10240,0.03124089042345683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,8192,0.026400888959566753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,12288,0.03130577670203315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,7168,0.0229751103454166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,10240,0.028326223293940227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,8192,0.0240639994541804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,7168,0.022296888960732356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,6144,0.020455999506844413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,5120,0.017785777648289997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,6144,0.02040177749262916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,4096,0.0148053334818946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,5120,0.018759111563364666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,3584,0.013257777525318993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,4096,0.017765333255132038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,3072,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,3584,0.017071111334694754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,3072,0.01625066664483812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,2560,0.010579555398888059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,2048,0.009380444056457942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,2560,0.015863999724388123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,1536,0.008029333419269985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,2048,0.014178666803571912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,1024,0.00572622236278322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,1536,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,768,0.004297777596447202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,1024,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,512,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,768,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,256,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,512,0.011634666886594562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,256,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,128,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,64,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,128,0.011176000038782755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,8192,32,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,64,0.011144888897736868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,8192,32,0.0120319997270902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,65536,0.15425066153208414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,51200,0.12360000610351562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,65536,0.10994311173756917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,16384,0.046652442879147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,51200,0.08537955416573419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,12288,0.036723554134368896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,16384,0.03575111097759671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,10240,0.03230844603644477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,12288,0.029527111185921565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,10240,0.026492445005310908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,8192,0.025058666865030926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,7168,0.021425777011447485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,8192,0.02260711126857334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,6144,0.01923466722170512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,7168,0.020809777908855014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,5120,0.016414221790101793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,6144,0.019621334142155118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,4096,0.013935110635227628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,5120,0.018295110927687753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,3584,0.012490666574902005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,3584,0.01663466625743442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,3072,0.011286222272449069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,3072,0.015651555524932016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,2560,0.009378666679064432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,2048,0.008553778131802877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,2560,0.015271999769740634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,1536,0.007140444384680853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,4096,0.017158221867349412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,1024,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,2048,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,768,0.004269333349333869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,1536,0.012861332959598966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,1024,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,512,0.003914666672547658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,256,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,768,0.012410666379663678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,512,0.01181333346499337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,128,0.0034888887570963967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,64,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,256,0.011156444748242697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,128,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,7168,32,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,64,0.011665778027640449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,65536,0.12653599845038518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,7168,32,0.01186044431395001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,51200,0.09959733486175537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,16384,0.03992977738380432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,65536,0.10057689083947076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,51200,0.07879110839631823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,12288,0.030942221482594807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,10240,0.027408000495698717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,12288,0.027225777506828308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,16384,0.03352088729540507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,8192,0.023771555887328252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,7168,0.01959733333852556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,10240,0.025176000263955858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,6144,0.017529777354664274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,7168,0.020006222857369315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,5120,0.015262222952312894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,6144,0.018941332896550495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,5120,0.017664889494578045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,4096,0.016826666063732572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,3584,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,3584,0.016343111793200176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,3072,0.010466666685210334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,8192,0.02113155523935954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,3072,0.015201777219772339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,2560,0.009422221945391761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,2560,0.0144195556640625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,2048,0.007811555431948767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,4096,0.012981333666377597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,1536,0.006313777632183499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,2048,0.01313244468635983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,1024,0.004604444321658877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,1536,0.013186666700575085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,768,0.003944888710975647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,1024,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,512,0.0036195553839206696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,768,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,256,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,512,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,128,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,256,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,128,0.0114720000161065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,64,0.003149333306484752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,6144,32,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,64,0.01111822244193819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,65536,0.10879111289978027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,6144,32,0.011914666328165265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,51200,0.08725688854853313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,65536,0.09178222550286187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,16384,0.03748355640305413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,51200,0.0732391079266866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,12288,0.02695377833313412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,16384,0.03155555658870273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,10240,0.02436977790461646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,12288,0.025273778372340735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,8192,0.02065955599149068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,10240,0.022407111194398668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,7168,0.018413333429230582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,8192,0.02015111181471083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,6144,0.015982222225930955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,7168,0.018981332580248516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,5120,0.013816888961527081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,6144,0.01828799976242913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,4096,0.012154666913880242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,4096,0.016302223006884258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,3584,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,3584,0.01573866605758667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,3072,0.00980088859796524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,3072,0.014158222410413953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,2560,0.008771555291281806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,5120,0.01722666621208191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,2048,0.007090667055712805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,2560,0.01347733371787601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,1536,0.0053582220441765254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,2048,0.012822222378518848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,1024,0.004263111286693149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,1536,0.012823999755912356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,768,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,1024,0.012139555480745105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,512,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,768,0.012240889171759287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,256,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,512,0.011572444604502784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,128,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,256,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,128,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,64,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,5120,32,0.0029813332690132987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,64,0.01075111081202825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,5120,32,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,65536,0.08815733591715495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,51200,0.07051022185219659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,65536,0.08599999878141616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,16384,0.028487109475665625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,51200,0.06828888919618395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,12288,0.023035556077957153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,16384,0.028368890285491943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,10240,0.02036533256371816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,12288,0.023074666659037273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,8192,0.01692800058258904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,10240,0.021408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,8192,0.019149333238601685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,7168,0.015772443678643968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,6144,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,7168,0.01808888879087236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,5120,0.011232888533009423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,4096,0.009511999785900116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,6144,0.01761866609255473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,3584,0.009087111386987898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,4096,0.014925332532988654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,5120,0.01648177703221639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,3072,0.008251555263996124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,2560,0.007605333295133378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,3584,0.01475822263293796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,2048,0.006249777972698212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,3072,0.013170666992664337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,1536,0.004600000050332812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,2560,0.01316177762216992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,1024,0.004173333446184794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,2048,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,1536,0.01239466667175293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,768,0.003952888978852166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,512,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,768,0.011807999677128263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,256,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,512,0.011811555259757571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,256,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,128,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,128,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,64,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,64,0.01038577738735411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,4096,32,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,1024,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,65536,0.07639466391669379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,4096,32,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,51200,0.06243999799092611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,65536,0.08262844218148126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,16384,0.026894221703211468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,51200,0.06684088706970215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,12288,0.02047911120785607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,16384,0.027461333407296076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,10240,0.018038221531444125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,12288,0.02197511163022783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,8192,0.01590133375591702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,10240,0.020803555846214294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,7168,0.014749333262443542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,8192,0.01827555563714769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,6144,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,7168,0.017976888351970248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,5120,0.010678222609890832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,6144,0.016920889417330425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,4096,0.009302222066455418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,5120,0.01609511176745097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,3584,0.00867555538813273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,4096,0.014174222946166992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,3072,0.008031999899281396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,3584,0.014277334014574686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,2560,0.00683111117945777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,3072,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,2048,0.005579555614127054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,2560,0.013209777573744455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,1536,0.004941333499219683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,2048,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,1024,0.004314666820896996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,1536,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,768,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,1024,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,512,0.0035822221802340615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,768,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,512,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,128,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,256,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,64,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,128,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3584,32,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,64,0.011038222246699862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,65536,0.06879288620418973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3584,32,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,51200,0.054935110939873584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,65536,0.07976710796356201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,16384,0.02306844459639655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,51200,0.06398044692145453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,12288,0.017792888813548617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,16384,0.02587822245226966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,10240,0.016156444946924847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,12288,0.021684444612926904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,8192,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,10240,0.02015644477473365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,7168,0.013832888669437833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,8192,0.018368888232443068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,6144,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,7168,0.017235555582576327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,5120,0.010303110712104374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,6144,0.016551110479566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,4096,0.008691555923885768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,5120,0.015436444017622205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,3584,0.008292444050312042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,4096,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,3072,0.00757244477669398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,3584,0.013809777796268463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,2560,0.006430222342411677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,3072,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,2048,0.0052488889131281115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,2560,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,1536,0.004892444445027245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,2048,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,1024,0.0042462220622433555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,1536,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,768,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,1024,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,512,0.0035662220584021676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,768,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,256,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,512,0.011365333365069495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,128,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,256,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,64,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,128,0.01071466671095954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,3072,32,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,64,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,65536,0.057856890890333385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,3072,32,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,51200,0.046691556771596275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,65536,0.07699111435148451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,16384,0.020376889242066275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,51200,0.062048892180124916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,16384,0.025031111306614343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,12288,0.01626488897535536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,12288,0.021203555994563635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,10240,0.015045333239767285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,10240,0.019547555181715224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,8192,0.013214222258991666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,8192,0.01791111131509145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,7168,0.01218577805492613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,7168,0.01724799970785777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,6144,0.011571555501884885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,6144,0.015860444969601102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,5120,0.00979200005531311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,5120,0.015040889382362366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,4096,0.008467555873923832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,4096,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,3584,0.007642666498819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,3584,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,3072,0.006963555183675554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,3072,0.013075555364290873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,2560,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,2560,0.012838222086429596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,2048,0.0052177777720822235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,2048,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,1536,0.004579555657174853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,1536,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,1024,0.0041848888827694785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,1024,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,768,0.003842666745185852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,768,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,512,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,512,0.01130488928821352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,256,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,256,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,128,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,128,0.010699555277824402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,64,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,64,0.010053333308961656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2560,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2560,32,0.011039111349317761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,65536,0.04905866583188375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,65536,0.07427733474307589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,51200,0.04010222355524699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,16384,0.017115554875797696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,51200,0.06011911233266195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,12288,0.015119999647140503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,16384,0.02470933397610982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,12288,0.021007110675175984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,10240,0.013736888766288757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,8192,0.01202933324707879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,10240,0.01939377850956387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,7168,0.011192888849311404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,8192,0.01813066667980618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,6144,0.0107360002067354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,7168,0.016210666961140104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,5120,0.00980355590581894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,5120,0.014557333456145393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,6144,0.015414221419228448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,4096,0.008692444198661381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,3584,0.00739644467830658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,4096,0.013863999810483722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,3072,0.006698666761318843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,3584,0.013584000368913015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,2560,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,3072,0.013261333107948303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,2560,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,2048,0.0052933332820733385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,1536,0.00463377767139011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,2048,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,1024,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,1536,0.012400888734393649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,1024,0.01184622198343277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,768,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,512,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,768,0.01182577759027481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,256,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,512,0.0118222220076455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,128,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,256,0.010984000232484607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,128,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,2048,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,64,0.010091555615266165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,65536,0.03895466526349386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,2048,32,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,51200,0.03129955463939243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,65536,0.07238844368192884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,16384,0.01516711049609714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,51200,0.058083554108937584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,16384,0.02401688860522376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,12288,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,10240,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,12288,0.020688888099458482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,8192,0.01051999959680769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,10240,0.019131556153297424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,7168,0.009875555833180746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,8192,0.016206221448050607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,6144,0.009120889008045197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,7168,0.015951999359660678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,5120,0.008674666285514832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,6144,0.014873777826627096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,5120,0.014402666025691561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,4096,0.00790666706032223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,4096,0.01349066694577535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,3584,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,3584,0.01351288871632682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,3072,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,3072,0.013192888763215808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,2560,0.005670222143332164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,2560,0.012848888834317526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,2048,0.0052471111218134565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,2048,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,1536,0.004609777695602841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,1536,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,1024,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,1024,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,768,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,768,0.011824888487656912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,512,0.00332622230052948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,512,0.011163555913501315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,256,0.0031679999083280563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,256,0.010847110715177325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,128,0.0029333333174387612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,128,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,64,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,64,0.010038221875826517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1536,32,0.0027342221389214196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1536,32,0.010795555180973478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,65536,0.027871112028757732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,65536,0.0699253347184923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,51200,0.02344088918632931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,51200,0.054937776592042714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,16384,0.01149511088927587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,16384,0.02369333306948344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,12288,0.010243555737866296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,12288,0.019125334090656705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,10240,0.010709332923094431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,8192,0.009297777381208207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,8192,0.016176889340082806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,7168,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,7168,0.015673778123325773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,6144,0.010016889207892949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,6144,0.014871110518773397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,5120,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,5120,0.014156444205178155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,4096,0.007361777954631382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,10240,0.017934223016103108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,4096,0.013548444542619916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,3584,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,3072,0.006305777778228124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,3584,0.013501333693663279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,2560,0.005653333332803514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,3072,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,2048,0.004940444396601783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,2560,0.012848888834317526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,1536,0.004642666627963384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,2048,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,1024,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,1536,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,768,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,1024,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,512,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,768,0.01178311142656538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,256,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,512,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,128,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,256,0.010796444283591377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,128,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,1024,32,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,64,0.010093332992659675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,65536,0.02511111067401038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,1024,32,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,51200,0.01995111174053616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,65536,0.06609955761167738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,16384,0.01092355532778634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,51200,0.05340177814165751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,12288,0.009474666582213508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,16384,0.02241244415442149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,10240,0.009382222261693742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,12288,0.019151111443837483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,8192,0.008390222158696916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,10240,0.017475555340449016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,7168,0.0081377774477005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,8192,0.016356445021099515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,6144,0.007429333196745978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,7168,0.015219555960761176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,5120,0.00796622203456031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,6144,0.014829332629839579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,4096,0.0074275558193524676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,3584,0.007288888924651676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,5120,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,3072,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,4096,0.01368711143732071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,2560,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,3584,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,3072,0.012817777693271637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,2048,0.00508622204264005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,2560,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,1536,0.004600000050332812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,2048,0.012136889000733694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,1024,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,1536,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,768,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,1024,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,512,0.0032604444358083936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,768,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,256,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,512,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,256,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,128,0.010381333529949188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,768,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,64,0.010804444551467896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,65536,0.017958222164048087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,768,32,0.01069155583779017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,51200,0.014933332800865173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,65536,0.06549600097868177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,16384,0.010423999693658618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,51200,0.05247377687030368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,12288,0.008671110702885522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,16384,0.022364444202846948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,10240,0.008700444466537898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,12288,0.018606222338146634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,8192,0.008054222497675154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,10240,0.01791466606987847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,7168,0.009024000002278222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,8192,0.01587377819750044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,6144,0.008365333080291748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,7168,0.015633778439627755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,5120,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,6144,0.014519110321998596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,4096,0.007262222468852997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,5120,0.014460444450378418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,3584,0.006614222294754452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,4096,0.013429332938459186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,3072,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,3584,0.013587555951542325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,2560,0.005589333259397083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,3072,0.01274222218328052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,2048,0.004942222187916438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,2560,0.013117333253224691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,2048,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,1536,0.004566222015354368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,1024,0.003932444585694207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,1536,0.012094222009181976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,768,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,768,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,512,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,512,0.011012444065676795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,256,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,128,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,128,0.010406222608354356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,1024,0.011728000309732227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,64,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,32,0.0026862221873468826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,512,32,0.011124444504578909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,65536,0.01484444406297472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,65536,0.06421777937147352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,51200,0.014539556370841132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,512,64,0.0029599999801980127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,16384,0.009259555902745988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,51200,0.0518142216735416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,12288,0.009377777576446533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,16384,0.021706667211320665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,10240,0.008634666601816813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,12288,0.019121777680185106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,8192,0.008034666379292807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,10240,0.017316444052590262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,7168,0.0087333329849773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,8192,0.01630577776167128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,6144,0.008366222182909647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,7168,0.015180443723996481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,5120,0.007644444704055786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,6144,0.01501066651609209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,4096,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,5120,0.014127110441525778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,3584,0.006614222294754452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,4096,0.013843555417325763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,3072,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,3584,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,2560,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,3072,0.01316177762216992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,2048,0.004898666507667965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,2560,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,1536,0.00424888895617591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,2048,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,1024,0.0035911111368073356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,1536,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,768,0.003510222252872255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,1024,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,512,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,768,0.011734222372372946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,256,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,512,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,128,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,256,0.01069866700304879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,64,0.002522666628162066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,128,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,64,0.01072266697883606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,256,32,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,256,32,0.01148088855875863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,65536,0.016584000653690763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,65536,0.06225688589943779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,51200,0.015145777000321282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,51200,0.04932711190647549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,16384,0.009007111191749573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,16384,0.02231377859910329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,12288,0.009320889082219865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,12288,0.019142222073343065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,10240,0.0085235552655326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,10240,0.01797155539194743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,8192,0.007730666961934831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,8192,0.016531555189026725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,7168,0.007636444436179266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,7168,0.015547555353906421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,6144,0.006959111326270633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,6144,0.014853333433469137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,5120,0.007489778101444244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,5120,0.014513777361975776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,4096,0.006698666761318843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,4096,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,3584,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,3584,0.01353422221210268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,3072,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,3072,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,2560,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,2560,0.013083555632167392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,2048,0.005603555589914322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,2048,0.012427555190192329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,1536,0.004304888761705822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,1536,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,1024,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,1024,0.011857777833938599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,768,0.0035128887328836652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,768,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,512,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,512,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,256,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,256,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,128,0.003120888852410846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,64,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,128,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,128,32,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,64,0.011408000356621213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,128,32,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,65536,0.01553600033124288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,51200,0.014449778530332776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,65536,0.05963644716474745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,16384,0.008339555727110969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,51200,0.04874133401446872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,12288,0.008015111088752747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,16384,0.022322666313913133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,10240,0.007312888900438945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,12288,0.019300444258583915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,8192,0.0069395556218094295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,10240,0.017873777283562552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,7168,0.006676444576846228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,8192,0.016562667157914903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,6144,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,7168,0.01555377741654714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,5120,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,6144,0.01520088811715444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,4096,0.006139555739031897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,5120,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,3584,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,4096,0.013855111267831592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,3072,0.005981333139869902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,3584,0.013840888937314352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,2560,0.005994666781690385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,3072,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,2560,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,2048,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,1536,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,1536,0.012451555165979596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,1024,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,1024,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,768,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,768,0.012182222472296821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,512,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,2048,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,256,0.00301155551440186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,512,0.011728888584507836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,128,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,64,0.0026862221873468826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,128,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,256,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,64,32,0.002532444480392668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,65536,0.015430221954981485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,32,0.011609777808189392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,65536,0.059432890680101186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,51200,0.014527999692493014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,16384,0.008716444174448649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,51200,0.04859822326236301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,12288,0.007962666451931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,16384,0.022389334109094407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,10240,0.007265778051482306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,12288,0.01926666663752662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,10240,0.017858665850427415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,8192,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,8192,0.0165964447789722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,64,64,0.01089244418674045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,7168,0.006656000183688269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,6144,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,7168,0.015497777197096082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,5120,0.006796444455782573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,6144,0.015222221612930298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,4096,0.0063475556671619415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,5120,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,3584,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,4096,0.01384622189733717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,3072,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,3584,0.013219555219014486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,3072,0.01313333296113544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,2560,0.005922666854328579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,2048,0.0052968888647026485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,2560,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,1536,0.004271999829345279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,2048,0.012083555261294046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,1024,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,1536,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,768,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,1024,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,512,0.002969777832428614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,768,0.012077333198653327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,256,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,512,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,256,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,128,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,64,0.0025635556214385563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,128,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,49,32,32,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,64,0.010750222537252637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,49,32,32,0.011504000259770287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,51200,0.5115093125237359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,51200,1.0637164645724826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,16384,0.3548559877607558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,16384,0.17262578010559082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,12288,0.26182132297092015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,12288,0.13346932993994817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,16384,0.2548746797773573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,10240,0.22071377436319986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,12288,0.19883999559614393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,10240,0.11470044983757867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,8192,0.17997333738538954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,8192,0.09273955557081436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,10240,0.1708284484015571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,7168,0.15744977527194554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,8192,0.14256355497572157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,7168,0.08328266938527425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,6144,0.13423111703660753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,51200,0.7405493524339465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,6144,0.07372177971733941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,5120,0.11320088969336616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,7168,0.12887467278374568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,5120,0.06359022193484835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,6144,0.1146355602476332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,4096,0.09268444114261204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,4096,0.05393155415852865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,5120,0.10121599833170573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,3584,0.04902933372391594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,4096,0.08741599983639187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,3072,0.07188000281651814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,3072,0.04424000117513868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,3584,0.07939200268851386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,2560,0.05932178099950155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,3584,0.08056533336639404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,3072,0.07336444324917264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,2560,0.03924977779388428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,2048,0.049730665153927274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,2048,0.03443466623624166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,1536,0.03854755560557047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,2560,0.08752089076571995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,2048,0.05983110931184557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,1536,0.02930755416552226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,1024,0.02776088813940684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,1536,0.0537431107627021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,1024,0.02300888962215847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,768,0.02107644412252638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,1024,0.0443102220694224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,512,0.015219555960761176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,768,0.019335110982259113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,768,0.04109244545300802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,256,0.009400000174840292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,512,0.016915554801623028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,512,0.034817778401904635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,128,0.0075697774688402815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,256,0.01516888870133294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,256,0.03135022189882066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,128,0.013802666631009845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,64,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,65536,32,0.006614222294754452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,65536,128,0.027466666367318895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,64,0.014723555909262763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,65536,32,0.017955554856194388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,65536,0.5440853436787924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,65536,1.0454951392279732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,51200,0.8173599772983127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,51200,0.4164248837365045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,16384,0.27573778894212514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,16384,0.1416364378399319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,16384,0.222126219007704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,12288,0.2031946712070041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,12288,0.11025244659847683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,65536,0.8090835677252876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,10240,0.17196178436279297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,12288,0.17343289322323271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,51200,0.638004461924235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,8192,0.14177066749996609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,10240,0.09512266847822402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,8192,0.07738577657275729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,7168,0.11938399738735622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,10240,0.150036440955268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,8192,0.12535556157430014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,7168,0.06983733177185059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,6144,0.10522400008307563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,6144,0.061920887894100614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,5120,0.08736799822913276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,7168,0.11425866021050347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,6144,0.10087733136283027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,5120,0.05420355664359199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,4096,0.0709386666615804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,5120,0.09061866336398655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,3584,0.06223377916547987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,4096,0.07829955551359388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,3584,0.04227911101447212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,3072,0.05402400096257528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,3584,0.07184888919194539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,3072,0.03841244512134128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,2560,0.04582488867971632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,3072,0.06595555278990003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,4096,0.046385778321160205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,2560,0.03431466552946303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,2048,0.038783109850353666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,2048,0.030003554291195337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,1536,0.02936799989806281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,2560,0.07502933343251546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,2048,0.05506755577193367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,1024,0.02075466679202186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,1536,0.025985777378082275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,1536,0.04845955636766222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,768,0.0162444445821974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,1024,0.02030311193731096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,1024,0.0411484440167745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,512,0.011811555259757571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,768,0.03774044579929776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,512,0.016526222229003906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,768,0.017924444542990792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,256,0.007816000117195977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,512,0.03440266516473558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,128,0.004604444321658877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,256,0.02896089024013943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,256,0.013399110900031196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,64,0.004252444538805219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,128,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,51200,128,0.02490844494766659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,51200,32,0.0042168887125121225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,64,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,51200,32,0.01694222291310628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,65536,0.3434826797909207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,65536,0.20000267028808594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,51200,0.259244441986084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,51200,0.15312355094485813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,16384,0.0875999993748135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,16384,0.05756622552871704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,65536,0.2214862240685357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,12288,0.06300889121161567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,16384,0.07301600111855401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,51200,0.17258044083913168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,10240,0.053658667537901134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,12288,0.060376889175838895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,10240,0.041336887412601046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,8192,0.04361066553327772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,10240,0.05478133426772224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,8192,0.0350213348865509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,7168,0.038506666819254555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,8192,0.04856977860132853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,7168,0.032283554474512734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,12288,0.046375112401114575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,6144,0.03407288922203912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,7168,0.045659555329216846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,6144,0.04249155521392822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,5120,0.029105779197480943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,5120,0.027032888597912256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,5120,0.03941244549221463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,4096,0.02370044423474206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,4096,0.02375999920898014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,4096,0.037279999918407865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,3584,0.021363554729355708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,3584,0.022051556242836848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,3072,0.018734221657117207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,3584,0.03425511055522495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,6144,0.029247111744350855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,2560,0.01591111057334476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,3072,0.020169778002632987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,3072,0.0328106681505839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,2048,0.013407111167907715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,2560,0.018939556346999276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,2560,0.03112444281578064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,2048,0.017603556315104168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,2048,0.030253334177864924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,1536,0.010850666297806634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,1536,0.016555555992656283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,1536,0.028540445698632136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,1024,0.008475555313958062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,1024,0.027083555857340496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,1024,0.013840888937314352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,768,0.00702133360836241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,768,0.025391111771265667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,512,0.005250666704442766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,512,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,256,0.004020444634887907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,512,0.022691556149058874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,256,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,128,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,256,0.02107644412252638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,128,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,64,0.003256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,16384,128,0.01969688965214623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,768,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,16384,32,0.003208000005947219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,64,0.011160000330872007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,16384,32,0.012425777812798819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,65536,0.17239999771118164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,51200,0.2021697759628296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,51200,0.13437867164611816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,65536,0.25172000461154515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,16384,0.0739004413286845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,65536,0.185296893119812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,16384,0.051220443513658315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,12288,0.051699555582470365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,16384,0.06435644626617432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,51200,0.1466924481921726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,12288,0.04165510998831855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,10240,0.044826666514078774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,12288,0.05388266510433621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,10240,0.03710133168432448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,8192,0.03766577773623996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,10240,0.04953866534762912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,7168,0.03193688723776076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,8192,0.03208533260557387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,7168,0.028967999749713477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,8192,0.044044444958368935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,6144,0.028447998894585505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,7168,0.04187822341918945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,5120,0.024392889605628118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,6144,0.03891377647717794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,6144,0.02597066594494714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,4096,0.020073778099483915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,5120,0.023825777901543513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,5120,0.03640177845954895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,3584,0.01793599956565433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,4096,0.021088000800874498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,4096,0.03472355670399136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,3072,0.016516443755891588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,3584,0.03257688879966736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,3072,0.019188443819681805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,2560,0.014387556248241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,3072,0.030890666776233252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,2560,0.017966222431924608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,2560,0.029856890439987183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,2048,0.012435555458068848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,2048,0.029446221060223047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,2048,0.017090666625234816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,1536,0.010240000155236986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,3584,0.01998666591114468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,1536,0.02846133377816942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,1024,0.008260444634490544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,1536,0.015442666080262927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,768,0.007063999772071838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,1024,0.023356444305843774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,1024,0.01368622233470281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,512,0.0046311111913787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,768,0.022691556149058874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,512,0.01236977759334776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,256,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,256,0.01181777815024058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,256,0.02225422196918064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,128,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,128,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,768,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,64,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,128,0.019537778364287484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,12288,512,0.02068800065252516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,12288,32,0.003289777785539627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,64,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,12288,32,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,65536,0.2120222250620524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,51200,0.16634666919708252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,65536,0.16494044992658827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,51200,0.1291946702533298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,16384,0.05736088752746582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,16384,0.04986844460169474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,65536,0.17063644197252062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,51200,0.13507110542721218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,12288,0.04159999887148539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,16384,0.061659554640452065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,10240,0.03525955478350321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,12288,0.051274667183558144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,12288,0.04049955474005805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,10240,0.03583555420239767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,8192,0.029394666353861492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,10240,0.04722222354676989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,8192,0.031050665511025324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,7168,0.025795555777019922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,8192,0.04217777649561564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,7168,0.02788177794880337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,6144,0.022760000493791368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,7168,0.040018667777379356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,6144,0.025411556164423626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,5120,0.019949333535300363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,6144,0.036877334117889404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,5120,0.022671111755900916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,5120,0.03533066643608941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,4096,0.015831111205948722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,4096,0.03262933426433139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,4096,0.020439111524158053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,3584,0.01422933406300015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,3584,0.031496001614464655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,3584,0.019283556275897555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,3072,0.012523555921183692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,3072,0.030020445585250854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,3072,0.0184906671444575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,2560,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,2560,0.02996266550487942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,2560,0.01758044461409251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,2048,0.010151111417346531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,2048,0.028548445966508653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,1536,0.008660444782839881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,2048,0.016484444340070088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,1536,0.013834666874673633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,1024,0.006816000160243776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,1536,0.027256889475716486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,1024,0.01277511070171992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,1024,0.023434667123688593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,768,0.004911111046870549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,768,0.021990221407678392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,768,0.012094222009181976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,512,0.0041982221106688184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,512,0.020619556307792664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,512,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,256,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,256,0.01886311173439026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,256,0.011355555719799466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,128,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,10240,128,0.017246221502621967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,128,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,64,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,10240,32,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,64,0.011041777829329172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,10240,32,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,65536,0.18046844005584717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,65536,0.1166986624399821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,51200,0.13534666432274714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,51200,0.09042221970028347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,16384,0.048376888036727905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,16384,0.03752711084153917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,65536,0.16375022464328343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,51200,0.13154844443003336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,16384,0.05840355820126004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,12288,0.03831200136078729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,10240,0.02846133377816942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,12288,0.03146933184729682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,12288,0.04982044299443563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,10240,0.02835822105407715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,8192,0.02352622151374817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,10240,0.04573422339227465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,8192,0.023046221998002794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,7168,0.021336888273557026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,8192,0.04115822248988681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,6144,0.018949333164427016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,7168,0.02161955502298143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,6144,0.020337777005301583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,7168,0.03945422172546387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,5120,0.01647200021478865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,6144,0.03671911027696397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,4096,0.012979555461141797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,5120,0.018925334016482036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,5120,0.034859554635153875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,3584,0.011922666596041786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,4096,0.01776266594727834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,4096,0.0328453348742591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,3584,0.017229333519935608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,3072,0.011492444409264458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,3584,0.03124177787039015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,2560,0.01036977767944336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,3072,0.016166221764352586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,3072,0.030533333619435627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,2048,0.008920888933870528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,2560,0.015837333268589444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,2560,0.029775109555986192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,1536,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,2048,0.014896889527638754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,2048,0.028965334097544353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,1024,0.005566222386227713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,1536,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,1024,0.012422222230169507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,1536,0.026729777455329895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,768,0.004910222358173794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,768,0.011887999872366587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,768,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,512,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,512,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,512,0.02148444453875224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,256,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,1024,0.022309333086013794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,256,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,128,0.003168888803985384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,256,0.018887110882335238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,128,0.011027555498811932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,8192,128,0.017550221747822232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,64,0.003031999907559819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,8192,32,0.002945777856641345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,64,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,8192,32,0.01203733351495531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,65536,0.15821154912312826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,65536,0.10848444700241089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,51200,0.11831200122833252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,51200,0.08384355571534897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,65536,0.16200266944037542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,16384,0.04511377877659268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,16384,0.03509955604871114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,51200,0.13053154945373535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,12288,0.035728000932269625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,16384,0.057652446958753795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,12288,0.029709332519107397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,10240,0.0313351121213701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,12288,0.0491466654671563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,10240,0.02645422187116411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,8192,0.02204888893498315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,10240,0.04553777641720242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,8192,0.021956443786621094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,8192,0.04073866539531284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,7168,0.019672888848516677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,7168,0.020627554919984605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,7168,0.03902933332655165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,6144,0.017254221770498488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,6144,0.019680889116393197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,6144,0.03629066546758016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,5120,0.01514400045077006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,5120,0.01827111177974277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,5120,0.03494400117132399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,4096,0.012455999851226807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,3584,0.011579555769761404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,4096,0.032152000400755144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,4096,0.0173342227935791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,3584,0.01651822196112739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,3072,0.010431110858917236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,3584,0.031121777163611516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,2560,0.009718221922715506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,3072,0.015916445189052157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,3072,0.03087466624048021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,2560,0.029247111744350855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,2048,0.009221333596441481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,2048,0.013215999636385175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,1536,0.007787555456161499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,2048,0.02826933397187127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,1536,0.012627555264366997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,1024,0.005906666732496685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,1536,0.0251226673523585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,1024,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,768,0.004332444320122401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,768,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,1024,0.023430221610599097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,2560,0.015385778413878547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,512,0.003943110919660992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,768,0.021941334009170532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,256,0.0032799999333090256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,512,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,512,0.022410665949185688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,256,0.010704889065689512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,256,0.019959999455346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,128,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,64,0.003015111097031169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,128,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,7168,128,0.01864177816443973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,7168,32,0.002942222274012036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,65536,0.12488533390892877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,32,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,7168,64,0.011815110842386881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,51200,0.09826578034294976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,65536,0.09974222050772773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,51200,0.07722933424843682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,65536,0.16107200251685247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,16384,0.03877866599294875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,51200,0.12989244196150038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,12288,0.03036266565322876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,16384,0.033482667472627424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,16384,0.05807110998365614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,12288,0.02666666607062022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,12288,0.048968887991375394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,10240,0.024152888192070857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,10240,0.045371555619769625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,8192,0.02286400066481696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,8192,0.020998222960366145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,8192,0.04008622301949395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,7168,0.019938665959570143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,7168,0.019661333825853135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,7168,0.038731555143992104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,6144,0.01795022189617157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,6144,0.018975110517607797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,10240,0.026736888620588515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,5120,0.015527110960748462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,6144,0.036248889234330915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,5120,0.018007111218240526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,5120,0.03497955534193251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,4096,0.01292711082432005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,4096,0.01685688893000285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,3584,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,4096,0.032660444577534996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,3584,0.016237333416938782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,3072,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,3584,0.03202133377393087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,3072,0.015531554818153381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,2560,0.009399111072222391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,3072,0.030644443300035264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,2560,0.013842666314707862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,2048,0.00755555596616533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,2560,0.029263112280103896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,2048,0.01315822203954061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,1536,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,2048,0.026163554853863184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,1536,0.012815111213260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,1024,0.004615111069546806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,1536,0.02516266703605652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,1024,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,768,0.004287999951177173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,1024,0.023327999644809302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,768,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,512,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,768,0.022710222336981032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,512,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,256,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,512,0.02232355541653103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,256,0.011530666715568967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,128,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,256,0.020999110407299463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,128,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,64,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,6144,128,0.01940711173746321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,6144,32,0.002955555501911375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,32,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,65536,0.10764000150892471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,65536,0.09107733435100979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,6144,64,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,51200,0.0862506694263882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,51200,0.07210666603512235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,65536,0.16018933720058864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,16384,0.036397334602144026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,16384,0.03128711051411099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,51200,0.12904355261060926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,12288,0.026160889201694067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,16384,0.05752888653013441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,12288,0.02507466740078396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,10240,0.0236817780468199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,12288,0.048918220731947154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,10240,0.044847998354170054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,8192,0.019652444455358718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,8192,0.020247111717859905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,8192,0.04045244389110141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,7168,0.017920888132519193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,10240,0.022487999664412603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,7168,0.038592000802357994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,7168,0.019296889503796894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,6144,0.016320889194806416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,6144,0.03618222143914964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,6144,0.01829777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,5120,0.014119111829333834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,5120,0.017217778497272067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,4096,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,4096,0.01621955633163452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,4096,0.032602667808532715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,3584,0.010573333336247338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,3584,0.01564088960488637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,5120,0.03464444478352865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,3584,0.031579554080963135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,3072,0.009388444324334463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,3072,0.014889778362380134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,3072,0.027479110492600337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,2560,0.008590222232871586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,2560,0.01349422252840466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,2560,0.027170666389995154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,2048,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,2048,0.026534222894244727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,2048,0.013265777793195514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,1536,0.005602666487296422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,1536,0.024404444628291663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,1536,0.012738666600651212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,1024,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,1024,0.023336889015303716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,1024,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,768,0.004204444587230682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,768,0.022727999422285292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,512,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,512,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,512,0.021306667062971327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,256,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,256,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,768,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,256,0.021293333835071985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,128,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,128,0.011499555574523078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,5120,128,0.019310222731696237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,64,0.011483555866612328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,5120,32,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,5120,32,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,65536,0.08682133091820611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,65536,0.08470933304892646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,65536,0.1600666708416409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,51200,0.07048800256517199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,51200,0.06796533531612821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,16384,0.029419554604424372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,51200,0.12865332762400308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,16384,0.02825333343611823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,12288,0.02162577708562215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,16384,0.0563031103875902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,10240,0.018754666050275166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,12288,0.022994667291641235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,12288,0.04797066582573784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,8192,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,10240,0.04458222124311659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,10240,0.02122666603989071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,7168,0.014807111687130399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,8192,0.040307554933759905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,7168,0.018139556050300598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,6144,0.01351111133893331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,7168,0.03797600004408095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,6144,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,5120,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,8192,0.01888799998495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,5120,0.016252444850073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,4096,0.009683555199040307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,5120,0.03430577781465318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,4096,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,4096,0.031101332770453557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,3584,0.008882666627566019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,3584,0.014496000276671516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,3072,0.008092444803979661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,3584,0.03089511063363817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,6144,0.03593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,2560,0.007350222104125553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,3072,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,3072,0.029262221521801416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,2048,0.006251555350091722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,2560,0.026879999372694228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,2560,0.01331733332739936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,1536,0.005024000174469418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,2048,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,2048,0.02511111067401038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,1024,0.003955555458863576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,1536,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,1536,0.024085332949956257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,1024,0.01180266671710544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,768,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,1024,0.023676445086797077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,768,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,512,0.003584889074166616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,512,0.02216800053914388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,512,0.011143999795118967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,256,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,256,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,256,0.02108088963561588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,128,0.0029333333174387612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,4096,128,0.018175111876593698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,128,0.011340444286664328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,64,0.002942222274012036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,4096,32,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,768,0.01145066652033064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,65536,0.07559733258353339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,64,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,4096,32,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,51200,0.061306668652428516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,65536,0.08227022488911946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,51200,0.06549777587254842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,65536,0.15997689300113255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,16384,0.025390222668647766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,51200,0.12923555903964573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,12288,0.01994311147265964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,16384,0.056623107857174344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,12288,0.02228533393806881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,10240,0.017084444562594097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,12288,0.04772888951831394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,10240,0.020502222908867728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,8192,0.01477688882086012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,10240,0.04394488864474826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,8192,0.018576888574494254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,7168,0.013880000346236758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,8192,0.0395760006374783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,7168,0.017640888690948486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,6144,0.01310400002532535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,7168,0.03800355394681295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,16384,0.026974222726292078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,5120,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,6144,0.017218665944205392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,6144,0.03492622243033515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,4096,0.0090515555606948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,5120,0.03397333290841844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,5120,0.01586933268441094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,4096,0.014906666345066495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,3584,0.008536888493431939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,4096,0.030596445004145306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,3584,0.014013333453072442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,3584,0.029440889755884807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,3072,0.00775111135509279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,3072,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,3072,0.027526222997241553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,2560,0.007074666519959767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,2560,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,2560,0.026781333817376032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,2048,0.005921777751710679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,2048,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,2048,0.024745777249336243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,1536,0.004892444445027245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,1536,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,1536,0.024749333659807842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,1024,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,1024,0.022416889667510986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,1024,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,768,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,768,0.011807999677128263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,512,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,512,0.011716444459226398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,512,0.02198666665289137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,256,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,256,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,768,0.022095110681321885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,128,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,256,0.020960888928837244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,128,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3584,128,0.01962222158908844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,64,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3584,32,0.0029120000286234748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,64,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3584,32,0.010909332997269101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,65536,0.06761955552630954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,65536,0.07821422153049044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,51200,0.05378489030732048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,65536,0.15981866253746882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,51200,0.06338489055633545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,16384,0.022282666630215112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,51200,0.12882222069634333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,12288,0.016788444585270353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,16384,0.026567111412684124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,12288,0.04758222235573662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,12288,0.021864889396561518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,10240,0.015523556205961438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,10240,0.044236444764667086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,10240,0.020051555501090158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,8192,0.013195555243227216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,8192,0.03984977801640829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,8192,0.018162666095627677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,7168,0.012519111235936483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,7168,0.0381279985109965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,16384,0.056573331356048584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,7168,0.01757422255145179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,6144,0.011125333607196808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,5120,0.010354667074150508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,6144,0.016740444633695815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,6144,0.03462666604253981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,5120,0.015618667006492615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,4096,0.008770667016506195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,4096,0.013816888961527081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,3584,0.008056888977686564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,4096,0.03091999888420105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,3584,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,3072,0.007682667010360294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,3584,0.02880355715751648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,3072,0.013156444662147097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,2560,0.0070168889231152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,3072,0.027100443840026855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,2560,0.012777778009573618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,2560,0.02601066728432973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,5120,0.03379822108480666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,2048,0.012509333590666452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,1536,0.004937777916590373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,2048,0.02571644385655721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,1536,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,1024,0.003923555629120933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,1536,0.024843555357721116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,1024,0.01183111137813992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,1024,0.02323466704951392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,768,0.00388355553150177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,2048,0.0059004442559348206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,512,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,768,0.0118568887313207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,768,0.02166844407717387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,256,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,512,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,512,0.022367111510700647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,256,0.019899555378490023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,128,0.002955555501911375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,256,0.011167111496130625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,64,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,128,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,3072,128,0.019394665956497192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,3072,32,0.0029724445194005966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,65536,0.061421334743499756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,64,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,3072,32,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,65536,0.07669422361585829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,51200,0.04719466633266873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,65536,0.16012267271677652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,51200,0.06130311224195692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,16384,0.020385776956876118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,51200,0.12910932964748806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,16384,0.025370667378107708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,12288,0.015043555034531487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,16384,0.05646222167544895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,12288,0.021609778205553692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,10240,0.014214222629865011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,12288,0.04843822121620178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,10240,0.019903111788961623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,10240,0.04434577624003092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,8192,0.012516444755925072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,8192,0.018266666266653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,8192,0.039656887451807656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,7168,0.011922666596041786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,7168,0.017529777354664274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,7168,0.037672887245814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,6144,0.010813333094120026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,6144,0.0165928883685006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,6144,0.03477422065205044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,5120,0.010075555907355415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,5120,0.015605333778593274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,5120,0.03232266505559286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,4096,0.008745777938101027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,4096,0.014125333891974555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,4096,0.02991466720898946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,3584,0.008064000142945183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,3584,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,3584,0.029471112622155085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,3072,0.007318221860461765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,3072,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,3072,0.027219555444187585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,2560,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,2560,0.02647200061215295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,2560,0.012783110969596438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,2048,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,2048,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,2048,0.025212445192866858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,1536,0.0047102222839991255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,1536,0.012448888685968188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,1536,0.02440799938307868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,1024,0.003968888686762916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,1024,0.023374223046832617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,1024,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,768,0.003924444317817688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,768,0.02274133265018463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,768,0.01348444405529234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,512,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,512,0.011148444480366178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,512,0.022034666604465906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,256,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,128,0.003293333368168937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,256,0.01145866678820716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,128,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,64,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,128,0.019288889235920377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2560,32,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,64,0.010071111222108206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,65536,0.04809866680039299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2560,32,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,65536,0.07447289095984565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,51200,0.039613333013322614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,51200,0.05923200315899319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,65536,0.15972267256842718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2560,256,0.0199991116921107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,16384,0.015216888652907478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,51200,0.1291333304511176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,16384,0.05641422006818983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,12288,0.015191111299726697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,16384,0.02480799953142802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,10240,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,12288,0.047360890441470675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,12288,0.021384888225131567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,8192,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,10240,0.01947111056910621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,10240,0.04383111000061035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,7168,0.011144888897736868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,8192,0.017960889471901786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,8192,0.039451556073294744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,6144,0.010642666783597736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,7168,0.01700444519519806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,7168,0.03591377867592706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,6144,0.03373333480623033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,5120,0.009995555712117089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,6144,0.015698666373888653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,4096,0.00812266684240765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,5120,0.014836443795098199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,5120,0.031927112076017596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,3584,0.007272889216740926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,4096,0.02882755464977688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,4096,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,3072,0.006734222173690796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,3584,0.027790221903059218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,3584,0.013774221969975365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,2560,0.006038222048017714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,3072,0.01312800000111262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,3072,0.0262106673585044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,2560,0.026756443911128577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,2560,0.012703110774358114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,2048,0.00526577772365676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,2048,0.02478222217824724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,1536,0.004615111069546806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,2048,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,1536,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,1536,0.024118221468395654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,1024,0.011815999945004782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,1024,0.022397332721286353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,768,0.0038053331275780997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,768,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,768,0.022015111313925848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,512,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,512,0.01148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,512,0.02100177771515316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,256,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,256,0.011027555498811932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,256,0.019596444235907663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,128,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,128,0.010727110836240979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,2048,128,0.01830400029818217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,64,0.00258933338854048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,64,0.010428444378905825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,32,0.002962666667169995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,2048,1024,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,2048,32,0.010817777779367236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,65536,0.03719110952483283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,65536,0.07175111108356051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,51200,0.03033511175049676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,65536,0.1598880026075575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,51200,0.05722755855984158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,16384,0.015105777316623263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,51200,0.128985775841607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,16384,0.02436622149414486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,16384,0.05554577708244324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,12288,0.012387555506494312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,12288,0.02070577773782942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,12288,0.046317332320743136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,10240,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,10240,0.04276177618238661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,10240,0.019147555033365887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,8192,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,8192,0.03831911087036133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,8192,0.016492444607946608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,7168,0.010455999937322406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,6144,0.010139555566840703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,7168,0.015917332635985482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,7168,0.03641866644223531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,6144,0.015042665931913586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,5120,0.009342222577995723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,6144,0.03363200028737386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,5120,0.014479110638300577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,4096,0.008376888930797577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,5120,0.03190222382545471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,3584,0.007060444189442529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,4096,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,4096,0.02955555584695604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,3072,0.006160888820886612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,3584,0.027426666683620874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,3584,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,2560,0.0059057776298787855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,3072,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,3072,0.02640977833006117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,2048,0.005229333208666907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,2560,0.01275199982855055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,2560,0.02600355611907111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,1536,0.004543111142185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,2048,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,2048,0.025791999366548326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,1024,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,1536,0.023382221659024555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,1024,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,768,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,1024,0.022301332818137273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,768,0.011702222128709158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,768,0.021535999245113794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,512,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,512,0.011016888750924004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,512,0.02129777769247691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,256,0.0031591111587153543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,256,0.018522666560279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,256,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,128,0.0029333333174387612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,128,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1536,128,0.01741333305835724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,64,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,1536,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1536,32,0.0025191110455327565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,65536,0.02659822172588772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,64,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1536,32,0.010709332923094431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,51200,0.022574222750133936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,65536,0.06941599978340997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,65536,0.15916089216868082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,16384,0.01165066659450531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,51200,0.05439555644989014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,51200,0.12858577569325766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,16384,0.023491554790072974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,12288,0.011156444748242697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,16384,0.055227554506725736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,12288,0.01909244391653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,12288,0.046075556013319224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,10240,0.01038044442733129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,10240,0.018218666315078735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,10240,0.042240000433391996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,8192,0.009751111268997192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,8192,0.016189333465364244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,7168,0.009124444590674506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,8192,0.03738311264250014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,7168,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,6144,0.00851999968290329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,7168,0.03566222058402167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,6144,0.014896889527638754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,6144,0.032280001375410296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,5120,0.008316444853941599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,5120,0.03162755568822225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,5120,0.014128888646761576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,4096,0.007344889144102733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,4096,0.028443555037180584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,4096,0.01351466692156262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,3584,0.006711999989218182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,3584,0.027463111612531874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,3584,0.013156444662147097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,3072,0.006261333409282897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,3072,0.02643555568324195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,3072,0.012843555874294706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,2560,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,2560,0.012458666331238218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,2560,0.02644177774588267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,2048,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,2048,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,1536,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,2048,0.024728889266649883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,1536,0.012033777932325998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,1536,0.024136000209384497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,1024,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,1024,0.022043555974960327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,768,0.0038399998512532976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,1024,0.01163200040658315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,768,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,512,0.003489777859714296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,768,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,256,0.0033271111961868075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,512,0.011008000208271874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,512,0.02104444470670488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,256,0.018599111172888014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,256,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,64,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,128,0.010944888823562197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,1024,128,0.01719377769364251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,1024,32,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,65536,0.024475556280877855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,32,0.010816000401973724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,65536,0.0658773316277398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,51200,0.0199617776605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,65536,0.15937688615587023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,51200,0.12830844190385607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,51200,0.05336533321274651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,16384,0.010954666468832227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,16384,0.05499022205670675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,16384,0.022847111026446026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,12288,0.010091555615266165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,1024,64,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,12288,0.04618311259481642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,12288,0.01902666687965393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,10240,0.01036266651418474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,8192,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,10240,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,10240,0.04220799936188591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,7168,0.008027555214034187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,8192,0.016554666890038382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,8192,0.03712888889842563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,6144,0.008645333349704742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,7168,0.035720000664393105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,7168,0.01550044450494978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,5120,0.007824888659848107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,6144,0.014827556080288358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,5120,0.03129511078198751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,5120,0.014064000712500678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,4096,0.007282666862010956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,4096,0.028492444091373022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,3584,0.006656888872385025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,4096,0.01349422252840466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,3584,0.013175999952687157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,3072,0.006269333263238271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,3584,0.0284826656182607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,3072,0.0127688886390792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,3072,0.02681955529583825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,2560,0.005943999936183293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,2560,0.026106667187478807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,2560,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,6144,0.03367200162675645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,2048,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,2048,0.026127111580636766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,1536,0.004605333424276776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,1536,0.024068444967269897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,1536,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,1024,0.003983111017280155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,1024,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,2048,0.005195555587609609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,768,0.0038337777886125776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,1024,0.023379556006855432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,768,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,512,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,768,0.022647110952271357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,512,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,256,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,512,0.022095110681321885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,256,0.010458666417333813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,128,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,256,0.02093511157565647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,128,0.010407999985747868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,768,128,0.01958400011062622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,64,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,768,32,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,64,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,768,32,0.010755555497275459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,65536,0.01778044468826718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,65536,0.06509155697292753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,51200,0.017215111189418368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,65536,0.15914311673906115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,51200,0.051944000853432544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,16384,0.010112000008424124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,51200,0.12856354978349474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,16384,0.02201955517133077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,12288,0.009702222214804756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,16384,0.05479200018776787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,12288,0.018598222070270114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,12288,0.04593511091338264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,10240,0.008959111240175035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,10240,0.041365332073635526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,10240,0.017900443739361234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,8192,0.007993777592976889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,8192,0.016553777787420485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,7168,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,7168,0.015895111693276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,7168,0.03568177753024631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,6144,0.008299555215570662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,6144,0.015122666954994202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,6144,0.03297422329584757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,8192,0.03678399986690945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,5120,0.007689778175618913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,5120,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,5120,0.03155733479393853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,4096,0.0069955554273393415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,4096,0.01385155568520228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,3584,0.006909333169460297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,3584,0.013545778062608508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,3584,0.02903199858135647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,3072,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,3072,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,3072,0.028135998381508723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,2560,0.005665777872006099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,4096,0.02849510974354214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,2560,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,2560,0.02626222206486596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,2048,0.004962666581074397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,2048,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,1536,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,2048,0.025457777910762366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,1536,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,1024,0.003912000192536248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,1536,0.024756444825066462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,1024,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,1024,0.02306933369901445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,768,0.003677333394686381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,512,0.0033928888539473214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,768,0.02163911031352149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,512,0.01995733380317688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,256,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,512,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,256,0.010332444475756751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,128,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,256,0.020688888099458482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,128,0.010404444403118556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,64,0.0025600000388092464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,512,128,0.019309333629078336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,512,32,0.002575111058023241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,64,0.00999111102686988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,65536,0.01529600057337019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,32,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,512,768,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,65536,0.06368355618582831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,51200,0.014878221684032016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,65536,0.15828711456722683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,16384,0.009401777552233802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,51200,0.05137066708670723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,16384,0.022050667140218947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,16384,0.05278311173121134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,12288,0.009383111364311641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,51200,0.12811733616722956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,10240,0.008777778181764815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,12288,0.018554665976100497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,12288,0.04554933309555054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,8192,0.007660444411966536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,10240,0.04120799899101257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,8192,0.016492444607946608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,10240,0.01791733337773217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,7168,0.007681777907742395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,8192,0.03740444448259141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,7168,0.01590399940808614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,7168,0.03502577874395583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,6144,0.0074471111098925276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,6144,0.01516888870133294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,5120,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,6144,0.032955557107925415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,5120,0.01443733274936676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,5120,0.031574222776624895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,4096,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,4096,0.013813333378897773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,3584,0.006585777633719974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,3584,0.013571555415789286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,3584,0.02918400035964118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,3072,0.006649777707126405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,3072,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,3072,0.02783999840418498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,2560,0.006317333214812809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,2560,0.012815111213260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,2560,0.02645955483118693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,2048,0.006439111298984951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,2048,0.012880889077981314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,4096,0.02986488739649455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,1536,0.004282666577233208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,2048,0.025771554973390367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,1536,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,1024,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,1536,0.024682665864626568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,1024,0.011920000116030375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,768,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,768,0.011683555940786997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,768,0.02279022170437707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,512,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,512,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,512,0.022389334109094407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,256,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,1024,0.023416888382699754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,256,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,128,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,256,0.020999999509917364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,128,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,256,128,0.01926488843229082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,256,32,0.0025315555847353404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,65536,0.01481244464715322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,64,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,256,32,0.011148444480366178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,51200,0.013541333377361298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,65536,0.06282755401399401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,51200,0.04895200000868904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,16384,0.008669333325492011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,65536,0.15845689508650038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,51200,0.12655377388000488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,16384,0.022127999199761286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,12288,0.007303111255168915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,16384,0.05387555559476217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,10240,0.007387555307812161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,12288,0.018952000472280715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,12288,0.04521244433191088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,8192,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,10240,0.041514668199751116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,10240,0.017856000198258292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,7168,0.006654222392373615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,8192,0.015967999895413715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,8192,0.03764177693261041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,6144,0.006325333482689328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,7168,0.015815110670195687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,7168,0.035394665267732404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,6144,0.01516355574131012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,5120,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,6144,0.033039109574423894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,4096,0.006602666858169768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,5120,0.014563555518786112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,5120,0.03100888927777608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,3584,0.00564444437623024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,4096,0.013780444032616086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,4096,0.029776000314288672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,3072,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,3584,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,3584,0.029233776860766943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,2560,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,3072,0.027149332894219294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,2560,0.012972444295883179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,2048,0.005952000204059813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,2560,0.02583822276857164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,2048,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,2048,0.025124443901909724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,1536,0.004582222137186262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,1536,0.02476177778508928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,1536,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,1024,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,1024,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,1024,0.02342222299840715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,768,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,768,0.02169599963559045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,512,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,768,0.01181422256761127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,512,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,512,0.02100444502300686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,256,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,3072,0.013181333740552267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,256,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,128,0.002560888934466574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,256,0.020605333977275424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,128,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,48,128,128,0.01874755488501655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,64,0.00258933338854048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,128,32,0.002513777878549364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,64,0.011041777829329172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,65536,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,128,32,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,51200,0.01037688884470198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,65536,0.060672885841793485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,16384,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,51200,0.04800533254941305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,12288,0.007363555332024892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,16384,0.022328888376553852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,10240,0.007252444823582967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,12288,0.018904889623324077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,8192,0.006619555668698416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,10240,0.017858665850427415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,7168,0.006578666468461354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,8192,0.015978667471143935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,7168,0.015807110402319167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,6144,0.005969777703285217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,5120,0.005971555494599872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,6144,0.015173332558737861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,4096,0.005581333405441708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,5120,0.014124444789356656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,3584,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,4096,0.013770666387346057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,3072,0.005594666633341048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,3584,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,2560,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,3072,0.013099555340078143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,2048,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,2560,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,1536,0.0042035554846127825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,2048,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,1024,0.003863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,1536,0.012085333466529846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,768,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,1024,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,512,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,768,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,256,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,512,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,256,0.010784889260927836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,64,0.002543111021320025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,128,0.011388444238238864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,64,32,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,64,0.010467555787828235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,64,32,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,65536,0.059566219647725425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,51200,0.0107342220014996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,51200,0.04821777674886915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,16384,0.007612444460391998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,16384,0.02198311189810435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,12288,0.007369777394665613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,12288,0.018909333480728995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,65536,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,10240,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,8192,0.006610666712125142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,10240,0.01756444407833947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,7168,0.006287999865081575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,8192,0.01651022169325087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,6144,0.005931555396980709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,5120,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,7168,0.01516888870133294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,6144,0.015120888749758402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,4096,0.00553955551650789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,3584,0.005535999933878581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,5120,0.014121777481502958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,3072,0.005579555614127054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,4096,0.013759999639458127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,2560,0.006118222243256039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,3584,0.01311555587583118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,2048,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,3072,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,1536,0.004276444514592488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,2560,0.012728888955381183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,1024,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,2048,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,768,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,1536,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,512,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,1024,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,256,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,768,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,128,0.002521777732504739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,512,0.011040888726711273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,64,0.0024888888001441956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,256,0.011342222491900126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,48,32,32,0.002519999941190084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,128,0.010416888528399997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,64,0.01034044474363327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,48,32,32,0.011272889044549731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,51200,0.505232016245524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,51200,1.0529830720689561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,16384,0.35628710852728945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,12288,0.26311911476982963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,16384,0.17049154970380995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,10240,0.21797066264682344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,12288,0.13200088342030844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,8192,0.18012977970971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,10240,0.11344799730512832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,7168,0.15456799666086832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,8192,0.09199111329184638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,7168,0.08276089032491048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,6144,0.13596799638536242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,5120,0.11245955361260308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,6144,0.07306310865614149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,4096,0.09383910894393921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,5120,0.06339733468161689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,3584,0.08023733562893338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,4096,0.05340266558859083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,3072,0.06950755251778497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,3584,0.0584462218814426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,2560,0.05914044380187988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,3072,0.04963199959860908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,2048,0.04896622233920627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,2560,0.04398577743106418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,2048,0.03919644488228692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,1536,0.037286221981048584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,1024,0.026491555902693007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,1536,0.031176000833511353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,768,0.02036711076895396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,1024,0.025472889343897503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,512,0.014110222458839417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,768,0.021670222282409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,256,0.009105777574910058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,512,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,128,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,256,0.014521777629852295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,64,0.006261333409282897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,128,0.014178666803571912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,65536,32,0.006639110959238476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,64,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,65536,32,0.015878222054905362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,65536,0.5356915791829427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,65536,1.0392915937635634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,16384,0.27802933586968315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,51200,0.8156719737582736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,51200,0.41132177246941465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,12288,0.2037644386291504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,16384,0.13962311214870876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,10240,0.17170133855607775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,12288,0.10870133505927192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,10240,0.09365689092212254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,7168,0.11869066291385227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,8192,0.14091911580827501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,7168,0.06913689110014173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,6144,0.104403555393219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,5120,0.08625600073072646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,6144,0.061328887939453125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,4096,0.07135644223954943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,5120,0.05371288789643181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,3584,0.061944888697730176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,4096,0.04574133290184868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,3072,0.05380444394217598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,3584,0.04907466636763679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,2560,0.04506577716933357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,3072,0.04303822252485487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,2048,0.0388648874229855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,8192,0.07663822174072266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,2560,0.037662221325768366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,1536,0.029197331931855943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,2048,0.03397066725624932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,1024,0.020690666304694284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,768,0.015971556305885315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,1024,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,512,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,768,0.019711110326978896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,512,0.016806221670574613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,256,0.007669332954618666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,128,0.004620444443490771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,256,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,64,0.004262222184075249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,64,0.014160888062583076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,51200,32,0.004228444563017951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,128,0.013463999662134381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,1536,0.027753776974148218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,51200,32,0.014548444085650973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,65536,0.3425920009613037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,51200,0.2652035554250081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,65536,0.18952533933851454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,16384,0.08559022347132365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,12288,0.06159200270970663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,16384,0.05605689022276136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,51200,0.14835911326938206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,10240,0.052314665582444936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,8192,0.04303466611438327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,12288,0.0453688899676005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,10240,0.040770666466818914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,7168,0.03773777683575948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,6144,0.033273776372273765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,8192,0.03458933366669549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,7168,0.0326204432381524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,5120,0.02832533253563775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,6144,0.029791110091739233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,4096,0.023154666026433308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,5120,0.027114666170544092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,3584,0.02091377807988061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,4096,0.023443554838498432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,3072,0.018618666463428073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,3584,0.025397333833906386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,2560,0.015674667225943673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,3072,0.02334044377009074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,2048,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,2560,0.021599110629823472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,2048,0.018955555227067735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,1536,0.010579555398888059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,1024,0.008382221890820397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,1536,0.017239999439981248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,768,0.006990222467316522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,1024,0.01513155632548862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,512,0.005062222066852781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,768,0.013830222189426422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,256,0.0035982223020659555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,512,0.012808889150619507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,128,0.0032506665835777917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,256,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,64,0.0032906666811969546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,128,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,64,0.012422222230169507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,16384,32,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,16384,32,0.012072000238630505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,65536,0.2510604328579373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,51200,0.19685067070855033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,65536,0.16586932871076795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,16384,0.07260800070232816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,51200,0.1293226612938775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,16384,0.05006755391756693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,12288,0.05332977904213799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,10240,0.04609333475430807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,12288,0.04103111227353414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,8192,0.03862577676773071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,10240,0.03637333379851447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,7168,0.03286666671435038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,8192,0.03180799881617228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,6144,0.029147555430730183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,7168,0.029808001385794744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,6144,0.026025777061780293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,5120,0.02511288887924618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,4096,0.020625778370433383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,5120,0.023185777995321486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,3584,0.018527110417683918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,4096,0.021041777398851182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,3584,0.02269511090384589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,3072,0.021360889077186584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,2560,0.014148443937301636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,2560,0.019320888651741874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,2048,0.012357333468066322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,2048,0.017645332548353408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,1536,0.010106667048401302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,3072,0.01643733349111345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,1536,0.016206221448050607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,1024,0.008120888637171851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,768,0.00674044465025266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,1024,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,512,0.004580444345871608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,768,0.013455111119482251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,512,0.012496888637542725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,256,0.0038782221575578055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,128,0.003622222277853224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,256,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,64,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,128,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,64,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,12288,32,0.0033973333322339584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,12288,32,0.011775111158688864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,65536,0.21978844536675346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,51200,0.1748755507998996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,65536,0.15807910760243735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,16384,0.06356000237994723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,51200,0.12456622388627793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,16384,0.04846577843030294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,12288,0.04680711030960083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,10240,0.035087111923429705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,12288,0.03940533267127143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,8192,0.029249777396519978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,10240,0.03557777735922072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,7168,0.025770665870772466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,8192,0.030693332354227703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,6144,0.02273066673013899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,7168,0.028558221128251817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,5120,0.019881778293185763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,6144,0.024604444702466328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,4096,0.016139555308553908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,5120,0.02278577784697215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,3584,0.014464888307783337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,4096,0.020649777518378366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,3072,0.012999999854299756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,3072,0.019504000743230183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,3584,0.02151288919978672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,2560,0.011398221883508893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,2560,0.018367111682891846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,2048,0.009975999593734741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,2048,0.01698666645420922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,1536,0.008433777425024245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,1536,0.015027556154463025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,1024,0.006743999818960826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,1024,0.013550221920013428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,768,0.004961777892377642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,768,0.013125333521101209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,512,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,256,0.0037360000941488477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,512,0.012207110722859701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,128,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,256,0.012049777640236748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,64,0.003177777760558658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,128,0.011144888897736868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,64,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,10240,32,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,10240,32,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,65536,0.17931200398339164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,51200,0.1342595550749037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,65536,0.11228622330559625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,16384,0.048637333843443126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,51200,0.0861448910501268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,12288,0.03722755445374383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,16384,0.03643733263015747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,10240,0.028520001305474177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,12288,0.030608889129426744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,8192,0.02363288899262746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,10240,0.027853333287768896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,7168,0.021134222547213238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,8192,0.02362755603260464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,6144,0.018668444620238412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,7168,0.022092445029152766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,5120,0.016424889365832012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,6144,0.02014755540423923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,4096,0.013200889031092325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,5120,0.018977777825461496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,3584,0.012141333685980903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,4096,0.017422222428851657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,3072,0.010823999842007955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,3584,0.018384888768196106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,2560,0.009863999982674917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,3072,0.01682133310370975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,2048,0.008617777791288164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,2560,0.016573333077960543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,1536,0.00739644467830658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,2048,0.014754666222466363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,1024,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,1536,0.013769778112570444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,768,0.004226666771703296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,1024,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,512,0.0036640001667870414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,768,0.012723555167516073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,256,0.0032337777730491427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,128,0.003032888803217146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,512,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,64,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,128,0.011036444041464063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,8192,32,0.0031537777847713898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,64,0.011011555790901184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,32,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,65536,0.15989777776930067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,65536,0.10021777947743733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,51200,0.11938044759962295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,16384,0.04436711139149136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,51200,0.07964799801508586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,12288,0.034761776526769005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,8192,256,0.01149866647190518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,16384,0.03457511133617825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,10240,0.03024622135692173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,8192,0.025555556019147236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,10240,0.025511110822359722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,7168,0.023192000057962205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,8192,0.02232711182700263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,6144,0.01921866668595208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,7168,0.020573332905769348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,5120,0.017229333519935608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,6144,0.019648889700571697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,4096,0.014064000712500678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,5120,0.018031110366185505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,4096,0.017151999804708693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,3584,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,3072,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,12288,0.02866311205757989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,2560,0.009296889106432596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,3584,0.017829333742459614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,2048,0.008092444803979661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,3072,0.016262221667501662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,2560,0.015248888068728976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,1536,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,1024,0.00480266660451889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,2048,0.014564444621404013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,1536,0.013156444662147097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,768,0.004206222378545337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,512,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,1024,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,768,0.012119111087587146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,256,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,128,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,512,0.011807111402352652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,256,0.011176000038782755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,64,0.0030871110243929755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,128,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,7168,32,0.00328177772462368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,64,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,65536,0.12248710791269939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,7168,32,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,65536,0.09177866909239028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,51200,0.07324000199635823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,16384,0.03770577907562256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,16384,0.03207822309599982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,12288,0.029574222034878198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,12288,0.02644088864326477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,10240,0.02586311101913452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,51200,0.09609778059853448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,8192,0.022078222698635522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,10240,0.02299733294381036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,7168,0.019674667053752475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,8192,0.021036444438828364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,7168,0.019715555840068393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,6144,0.017567111386193168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,5120,0.015882665912310284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,6144,0.01850755512714386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,4096,0.011590222517649332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,5120,0.01779733267095354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,3584,0.01180266671710544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,4096,0.01627466744846768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,3072,0.010495999621020423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,3584,0.01695999999841054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,3072,0.01588000026014116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,2560,0.009649777577983009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,2048,0.007672888537247975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,2560,0.014556444353527494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,1536,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,2048,0.013854222165213691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,1024,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,1536,0.012843555874294706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,768,0.0041555555330382455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,1024,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,768,0.011789333489206104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,512,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,256,0.0032559999575217566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,256,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,128,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,128,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,64,0.002982222164670626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,64,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,512,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,6144,32,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,65536,0.1050986647605896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,6144,32,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,51200,0.0835955540339152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,65536,0.08446577522489761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,16384,0.035586665074030556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,51200,0.06792000267240736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,12288,0.02661955522166358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,16384,0.029855112234751385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,10240,0.023054222265879314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,8192,0.020030222005314298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,10240,0.02198133369286855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,8192,0.01956622302532196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,7168,0.017627555463049147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,6144,0.0159288893143336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,7168,0.018591110905011494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,5120,0.014567999376191033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,6144,0.017978666557206046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,4096,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,5120,0.016682666209008958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,3584,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,4096,0.015917332635985482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,3072,0.009722666607962714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,3584,0.016240888171725802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,2560,0.008046222229798635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,3072,0.014427555931939019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,2048,0.0069982219073507525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,2560,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,1536,0.00555377784702513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,12288,0.02515022291077508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,2048,0.013190222283204397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,1024,0.004616888860861461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,1536,0.012446222205956778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,768,0.003979555434650845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,1024,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,512,0.0035991109907627106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,768,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,256,0.00332266671790017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,512,0.011710222396585675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,128,0.0032586666444937387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,256,0.011041777829329172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,64,0.0029688889367712867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,128,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,5120,32,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,64,0.010432889064153036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,5120,32,0.011125333607196808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,65536,0.0847582221031189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,65536,0.07709422376420763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,51200,0.06848711437649198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,16384,0.028991109795040552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,12288,0.02148088812828064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,16384,0.027207111318906147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,51200,0.06358666552437676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,12288,0.022314666046036616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,10240,0.020777778493033517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,8192,0.015903110305468243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,8192,0.018927110566033255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,7168,0.014877332581414117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,7168,0.01792711185084449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,6144,0.012492444780137805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,10240,0.01941511034965515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,5120,0.01126844435930252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,6144,0.017143999536832173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,4096,0.009398221969604492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,5120,0.016512889001104567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,4096,0.015184889237085978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,3584,0.008922666311264038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,3072,0.008064000142945183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,3584,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,2560,0.007326222128338284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,2560,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,3072,0.013778666655222574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,2048,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,1536,0.004909333255555895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,1536,0.012663111090660095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,1024,0.004267555558019214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,2048,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,768,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,1024,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,512,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,768,0.011815110842386881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,256,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,256,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,512,0.01181244436237547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,128,0.002959111084540685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,64,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,64,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,4096,32,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,128,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,65536,0.07383910814921062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,4096,32,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,51200,0.059598220719231494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,65536,0.07527022229300605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,16384,0.02685333291689555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,16384,0.025536888175540503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,12288,0.020480889413091872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,12288,0.022076444493399724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,51200,0.06107822391721937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,10240,0.02035911050107744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,8192,0.014685332775115967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,8192,0.018831999765502084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,7168,0.014067555467287699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,7168,0.017858665850427415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,6144,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,10240,0.018383999665578205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,6144,0.016689777374267578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,5120,0.01216533366176817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,4096,0.009287111461162567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,5120,0.01595288846227858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,3584,0.008423999779754216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,3584,0.014168888330459595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,4096,0.014555555250909595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,3072,0.007954667011896769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,2560,0.006980444822046492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,2560,0.01310400002532535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,3072,0.013545778062608508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,2048,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,1536,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,1536,0.012514666550689273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,2048,0.012841777669058906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,1024,0.012127999630239276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,768,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,768,0.012176888684431711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,512,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,256,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,1024,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,512,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,128,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,256,0.010772444307804108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3584,32,0.00299466670387321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,128,0.011496889094511667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,64,0.011085333095656501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,65536,0.06628621949089898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3584,32,0.010505778094132742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,51200,0.0524088872803582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,65536,0.07240621911154853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,16384,0.02196177840232849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,12288,0.0170462214284473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,51200,0.05904889106750488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,16384,0.0247706671555837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,10240,0.01571466690964169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,8192,0.013236444857385425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,10240,0.01962222158908844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,12288,0.021406221720907424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,7168,0.012606222596433429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,6144,0.011493333511882357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,8192,0.017989334132936265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,5120,0.010218666659461128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,7168,0.01756888959142897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,6144,0.016813332835833233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,4096,0.008937777744399177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,5120,0.015379556351237826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,3584,0.008093333078755273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,4096,0.014093332820468478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,3072,0.007371555599901412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,3584,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,2560,0.006774222271309958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,3072,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,2048,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,2560,0.013210666676362356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,1536,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,2048,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,1024,0.004219555606444676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,1536,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,768,0.0036915557252036203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,1024,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,512,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,768,0.01183377785815133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,256,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,512,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,256,0.010757333702511258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,64,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,128,0.010407999985747868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,3072,32,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,64,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,65536,0.06051377455393473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,3072,32,0.010438222024175854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,51200,0.045034666856129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,65536,0.06982488764656915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,16384,0.02020355562369029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,51200,0.056982219219207764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,12288,0.01551644504070282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,16384,0.02416177756256527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,10240,0.01405066748460134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,12288,0.020636444290479023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,8192,0.012543111211723752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,10240,0.01972888906796773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,7168,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,8192,0.017869333426157635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,6144,0.01075644459989336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,7168,0.01723911033736335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,5120,0.00978577799267239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,6144,0.015781333049138386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,4096,0.008807111117574904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,5120,0.01518133282661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,3584,0.008173333273993598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,4096,0.013578666581047906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,3072,0.0069511110583941145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,3584,0.013752000199423896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,2560,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,3072,0.012851555314328937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,2048,0.005579555614127054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,2560,0.013103110922707452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,1536,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,2048,0.012475555141766867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,1024,0.004263111286693149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,768,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,1536,0.012299555871221753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,1024,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,512,0.0035582222044467926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,256,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,768,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,128,0.003167111012670729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,512,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,256,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2560,32,0.002966222249799304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,128,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,64,0.010409778190983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,65536,0.0488666660255856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,51200,0.04013155566321479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2560,32,0.010481778118345471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,65536,0.06731910838021173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,16384,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,12288,0.014884443746672736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,51200,0.05531733234723409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,16384,0.02347733411524031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,10240,0.013542222479979197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,8192,0.01201600001917945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,12288,0.02072266737620036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,10240,0.018937778141763475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,7168,0.011163555913501315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,6144,0.010417777631017896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,8192,0.017695110705163743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,7168,0.01622577839427524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,5120,0.00999466660949919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,4096,0.00847466703918245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,6144,0.015474667151769003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,3584,0.0074177781740824384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,5120,0.014474666780895658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,3072,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,4096,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,3584,0.013842666314707862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,2560,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,3072,0.01278222186697854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,2048,0.005255110975768831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,1536,0.004648888690604104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,2048,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,2560,0.012808000048001608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,1024,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,1536,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,768,0.003825777934657203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,1024,0.011417778001891242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,512,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,768,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,512,0.011491555306646558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,128,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,256,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,64,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,128,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,2048,32,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,64,0.010307555397351583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,65536,0.0366977784368727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,2048,32,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,51200,0.030105776256985132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,65536,0.065482669406467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,16384,0.015024888846609326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,51200,0.053080889913770885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,16384,0.02306844459639655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,12288,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,10240,0.011576889289749993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,12288,0.019985778464211356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,8192,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,10240,0.018574222922325134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,8192,0.01624000072479248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,7168,0.010141332944234213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,6144,0.0105262224872907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,7168,0.015463110473420886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,5120,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,6144,0.015211555692884656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,4096,0.008022222254011366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,5120,0.014087110757827759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,3584,0.007284444239404466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,4096,0.013832888669437833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,3584,0.013549333645237817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,3072,0.00638488887084855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,2560,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,3072,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,2048,0.004936888813972473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,2560,0.013094222380055321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,1536,0.0046302220887608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,2048,0.012154666913880242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,1024,0.0042239998777707415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,1536,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,1024,0.01148977792925305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,768,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,512,0.0035004446076022256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,768,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,512,0.011508444117175208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,128,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,256,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,64,0.002806222273243798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,64,0.01036888857682546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,128,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1536,32,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,65536,0.026389333936903212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1536,32,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,51200,0.022466666168636743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,65536,0.0625457763671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,16384,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,51200,0.04979822370741102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,16384,0.022051556242836848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,12288,0.010223999619483948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,10240,0.01037599974208408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,12288,0.018595554762416415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,8192,0.00906488878859414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,10240,0.017872888180944655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,7168,0.008354666332403818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,8192,0.0158906661801868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,6144,0.007676444119877285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,7168,0.015836444165971544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,6144,0.014911999305089315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,5120,0.008303110798199972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,4096,0.007481777833567725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,5120,0.014140443669425117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,3584,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,4096,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,3072,0.006320888797442119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,3584,0.01351199961370892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,2560,0.005607111172543631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,3072,0.01316177762216992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,2560,0.01276266657643848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,2048,0.005224888937340842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,1536,0.004606222112973531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,2048,0.012404444317022959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,1024,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,1536,0.012418666647540199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,768,0.003955555458863576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,1024,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,512,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,768,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,512,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,256,0.003169777699642711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,128,0.003168888803985384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,256,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,128,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,64,0.01037599974208408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,1024,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,1024,32,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,65536,0.024850666522979736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,65536,0.06059110826916165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,51200,0.019861333900027804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,16384,0.010789333118332757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,51200,0.04902044600910611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,12288,0.009447111023796929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,16384,0.02148355543613434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,12288,0.018260444204012554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,10240,0.009848888549539778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,8192,0.008752889103359645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,10240,0.016954667038387723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,7168,0.009728888670603434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,8192,0.016203555795881484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,6144,0.008355555435021719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,7168,0.015223110715548197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,5120,0.007939555578761632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,6144,0.01482399966981676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,4096,0.007279111279381647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,5120,0.013819555441538492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,3584,0.006912888752089606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,4096,0.013510222236315409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,3072,0.005998222364319696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,3584,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,2560,0.005615111026499007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,3072,0.01276622215906779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,2048,0.004899555610285865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,2560,0.012401777837011548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,1536,0.004604444321658877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,2048,0.012410666379663678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,1024,0.0038542221817705366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,1536,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,768,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,1024,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,512,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,768,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,256,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,512,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,128,0.0028435554769304064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,256,0.010677333507272931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,64,0.002512000087234709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,128,0.011042666931947073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,768,32,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,64,0.010037333601050908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,65536,0.01756977703836229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,768,32,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,51200,0.014716444744004143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,65536,0.05925777885648939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,16384,0.010057777994208867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,51200,0.04893955588340759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,12288,0.01036088913679123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,16384,0.02073955535888672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,10240,0.009344889058007134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,12288,0.018559111489189994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,8192,0.008309333688682979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,10240,0.016869333055284288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,7168,0.007684444387753804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,8192,0.016234666109085083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,6144,0.008346666892369589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,7168,0.01512888901763492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,5120,0.007623111208279927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,6144,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,4096,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,5120,0.013801777528391944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,3584,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,4096,0.013415999710559845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,3072,0.006275555739800136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,3584,0.013447999954223633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,2560,0.005607111172543631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,3072,0.012734221915404001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,2048,0.004888888862397936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,2560,0.012469333079126147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,1536,0.004554666578769684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,2048,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,1024,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,1536,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,768,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,1024,0.011657777759763928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,512,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,512,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,256,0.0029360000044107437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,256,0.011365333365069495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,128,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,128,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,64,0.0025048889219760895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,64,0.011128889189826118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,512,32,0.0028328889360030494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,768,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,65536,0.015110222829712762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,512,32,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,51200,0.014341332846217685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,65536,0.05829777982499865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,51200,0.04591644472546048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,16384,0.021329777108298406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,12288,0.009263111485375298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,12288,0.01861422260602315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,10240,0.00871733327706655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,10240,0.01717066599263085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,16384,0.009395555489593083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,8192,0.008022222254011366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,8192,0.015528000063366361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,7168,0.008642666869693333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,7168,0.014861333701345654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,6144,0.008016889293988546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,6144,0.014479999740918478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,5120,0.0076142218377855085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,5120,0.013776000175211163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,4096,0.0069546666410234236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,4096,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,3584,0.006576889091067844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,3584,0.013506666653686099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,3072,0.006568000134494569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,3072,0.012668444050682915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,2560,0.006256888724035687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,2560,0.012455999851226807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,2048,0.004929777648713854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,2048,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,1536,0.004513777792453766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,1536,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,1024,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,1024,0.011016888750924004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,768,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,768,0.011339555184046427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,512,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,512,0.01111644423670239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,256,0.003148444410827425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,256,0.010657777388890585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,128,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,128,0.010393777655230628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,64,0.002529777793420686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,64,0.010050666828950247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,256,32,0.002579555536309878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,256,32,0.010076444182131026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,65536,0.016013332539134555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,65536,0.05483822358979119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,51200,0.014874666929244995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,16384,0.009014222357008193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,16384,0.020634666085243225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,12288,0.007962666451931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,51200,0.04472088813781738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,10240,0.00886044485701455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,12288,0.017941332525677152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,8192,0.007613333563009898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,7168,0.007617777420414819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,10240,0.017151110702090792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,8192,0.015610666738616096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,6144,0.006969778074158563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,5120,0.006575999988449945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,7168,0.015882665912310284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,6144,0.014470222923490735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,4096,0.0063440000845326324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,3584,0.006235555642180973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,4096,0.013359111216333179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,3072,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,3584,0.013829333086808523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,3072,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,2560,0.006250666661394968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,2048,0.0059262220230367445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,2560,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,1536,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,2048,0.01275288893116845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,1024,0.0038968887594011095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,1536,0.011900444825490316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,768,0.003491555651028951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,1024,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,512,0.003363555504216088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,5120,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,256,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,512,0.011765333513418833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,128,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,256,0.010701333483060202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,64,0.002516444358560774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,128,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,128,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,64,0.010144000252087912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,65536,0.01513777838812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,32,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,51200,0.013220444321632385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,128,768,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,65536,0.05432088838683235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,16384,0.007981333467695449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,12288,0.007359111474619971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,51200,0.044857776827282377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,10240,0.0070159998204973005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,16384,0.021014221840434607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,12288,0.017916444275114272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,8192,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,7168,0.006360000206364526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,10240,0.01757422255145179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,8192,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,6144,0.0063946665161185795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,5120,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,7168,0.015169777803950839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,6144,0.014197332991494073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,4096,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,3584,0.005971555494599872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,5120,0.014527110589875115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,4096,0.013530666629473368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,3072,0.00573600000805325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,3584,0.01349866638580958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,2560,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,3072,0.013210666676362356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,2048,0.005962666538026597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,2560,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,1536,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,2048,0.012528888881206512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,1024,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,1536,0.012148444851239523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,768,0.003583999971548716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,1024,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,512,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,768,0.011734222372372946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,256,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,512,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,128,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,256,0.010767111347781287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,64,0.0027742222365405825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,128,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,64,32,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,64,0.010088000032636855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,64,32,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,65536,0.015242666006088257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,51200,0.012835555606418185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,65536,0.05447466505898369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,16384,0.00815822184085846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,51200,0.04463911056518555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,12288,0.007357333269384172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,16384,0.02129777769247691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,12288,0.01824355622132619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,10240,0.007098666495747036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,8192,0.006648889018429651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,10240,0.01696977847152286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,8192,0.0162435554795795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,7168,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,6144,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,7168,0.015795555379655626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,5120,0.005944889038801193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,6144,0.014675555957688225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,4096,0.005612444546487596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,5120,0.014462222655614218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,3584,0.006004444426960415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,4096,0.013507555756303998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,3072,0.005650666852792104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,3584,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,2560,0.006305777778228124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,3072,0.012816888590653738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,2048,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,2560,0.01313688854376475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,1536,0.004584889031118817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,2048,0.0121742222044203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,1024,0.003973333372010125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,1536,0.012452444268597497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,768,0.0036115555299652945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,1024,0.011536888778209686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,512,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,768,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,256,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,512,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,128,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,256,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,64,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,128,0.010465777582592435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,33,32,32,0.002551111082235972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,64,0.011356444822417365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,33,32,32,0.010247111320495605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,51200,0.5055271254645454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,51200,1.0634062025282118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,16384,0.35527645217047793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,16384,0.17079199684990776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,12288,0.2623902161916097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,12288,0.13198755847083196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,16384,0.25230932235717773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,10240,0.21848977936638725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,12288,0.19767644670274523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,10240,0.11373955673641628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,8192,0.17672533459133574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,8192,0.09212888611687554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,10240,0.16985600524478486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,7168,0.15523733033074275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,8192,0.14142044385274252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,7168,0.08289155695173475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,6144,0.13599199718899196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,6144,0.07314488622877333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,51200,0.7294764518737793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,5120,0.11204711596171062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,7168,0.12833600574069554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,5120,0.06341422266430326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,4096,0.09222133292092217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,6144,0.11463199721442328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,4096,0.053468445936838783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,5120,0.10002577967113918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,3584,0.08035822047127618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,3584,0.05243733194139269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,4096,0.08658399846818711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,3072,0.06812089019351535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,3072,0.049534221490224205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,3584,0.08624000019497341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,2560,0.05739644500944349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,2560,0.04334311021698845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,3072,0.07634311252170138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,2048,0.048115554783079356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,2560,0.09179111321767171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,2048,0.03595644566747878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,1536,0.03660710983806186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,2048,0.06532355811860827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,1536,0.02952444553375244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,1024,0.02567022211021847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,1536,0.05232622226079305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,1024,0.02256444427702162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,1024,0.045313778850767344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,768,0.019620445039537217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,768,0.038943999343448214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,512,0.014051554931534661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,512,0.01811555524667104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,512,0.03527733352449205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,256,0.008649778034951951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,768,0.020311999652120803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,128,0.006628444625271692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,256,0.031719999180899724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,256,0.014947555131382413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,64,0.005966222120655908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,128,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,65536,32,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,64,0.015286222100257874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,65536,32,0.015503111812803479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,65536,0.5362008942498101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,65536,128,0.026178666286998328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,65536,1.0334284040662978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,51200,0.4104986720614963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,51200,0.8002008861965604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,16384,0.27322223451402455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,16384,0.13971911536322698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,16384,0.2201697826385498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,12288,0.19919555717044404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,12288,0.10847910907533433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,12288,0.17203377352820504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,51200,0.629093329111735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,10240,0.16663200325436062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,10240,0.0936808917257521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,65536,0.7966968748304578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,8192,0.1384062237209744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,8192,0.07685066594017877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,10240,0.1486942238277859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,7168,0.11916889084710015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,7168,0.06907111406326294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,8192,0.12451644738515218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,6144,0.10265422529644436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,6144,0.06121155288484362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,5120,0.08652889066272312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,5120,0.05352266629536947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,7168,0.11284622881147598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,6144,0.10167466931872898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,4096,0.07329066594441731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,4096,0.045463999112447105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,3584,0.06177600224812826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,5120,0.08912266625298394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,4096,0.0767680009206136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,3072,0.05475644601715935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,3584,0.04471288786994087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,3584,0.0848311119609409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,2560,0.04639644424120585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,3072,0.06798222329881456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,2560,0.035234666532940336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,2048,0.03917600048912896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,2048,0.03090933296415541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,2560,0.0792906681696574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,1536,0.029378665818108454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,2048,0.058182219664255776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,1536,0.026179555389616225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,1024,0.02107200026512146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,1536,0.04795111219088236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,1024,0.01997866729895274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,768,0.016013332539134555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,1024,0.04002933369742499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,3072,0.03982933362325033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,768,0.01796800063716041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,512,0.011863999896579318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,768,0.03548177745607164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,256,0.007872000336647034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,512,0.016708445217874315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,512,0.03259199857711792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,256,0.013847110999955071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,256,0.029174221886528864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,128,0.004566222015354368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,64,0.003947555604908201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,128,0.01313866674900055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,51200,128,0.02507200009293026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,51200,32,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,32,0.014851555228233337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,51200,64,0.014122666584120857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,65536,0.3362390995025635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,65536,0.18969155682457817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,51200,0.24889511532253691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,51200,0.14830756187438965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,16384,0.0872231125831604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,16384,0.05605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,16384,0.07107110818227132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,12288,0.06141599681642321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,51200,0.16505244043138292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,65536,0.21101421780056426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,12288,0.04560533497068617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,10240,0.05261422197024027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,12288,0.059192001819610596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,10240,0.04054933455255296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,8192,0.04302488764127096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,10240,0.053716444306903414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,8192,0.03472977876663208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,7168,0.03774044579929776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,8192,0.04719110992219713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,7168,0.03189599845144484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,6144,0.033054222663243614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,7168,0.04496711161401537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,6144,0.028957333829667833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,5120,0.028413332170910303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,5120,0.02642400066057841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,6144,0.04265422291225857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,5120,0.03877777854601542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,4096,0.023019555542204116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,3584,0.020787555310461257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,4096,0.0348053342766232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,4096,0.022787556052207947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,3584,0.0233608881632487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,3072,0.01832444469134013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,3584,0.033519112401538424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,3072,0.021706667211320665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,2560,0.015545777148670621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,3072,0.03237511052025689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,2048,0.012960000170601739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,2560,0.019660444723235238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,2560,0.031055112679799397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,1536,0.010203555226325989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,2048,0.029351999362309773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,2048,0.018208889497650992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,1024,0.008198222352398766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,1536,0.02760266595416599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,1536,0.016509332590632968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,768,0.007281777759393056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,1024,0.014171555638313293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,768,0.024991111622916326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,768,0.012517333030700684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,512,0.004911111046870549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,512,0.022036444809701707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,512,0.012168889244397482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,256,0.003591999825504091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,256,0.021018667353524104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,256,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,128,0.0031475555151700974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,1024,0.02511466708448198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,64,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,16384,128,0.01942044496536255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,128,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,16384,32,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,64,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,16384,32,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,65536,0.24664711952209473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,65536,0.16567199760013157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,51200,0.19234489070044625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,51200,0.12877599398295084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,16384,0.0726755526330736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,16384,0.04984177814589607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,65536,0.17550399568345812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,16384,0.06291377544403076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,51200,0.13891022735171848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,12288,0.05286844571431478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,10240,0.046078221665488354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,12288,0.04057333204481337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,12288,0.05338222119543287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,10240,0.03657866517702738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,8192,0.03814399904674954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,10240,0.0484186675813463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,7168,0.03273511264059279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,8192,0.03164888752831353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,8192,0.043396443128585815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,7168,0.02905511193805271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,6144,0.029489778810077246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,7168,0.04100355505943298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,6144,0.026367111338509455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,6144,0.03860977623197768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,5120,0.023503111468421087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,4096,0.020642666353119742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,5120,0.03552711009979248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,4096,0.02088177700837453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,4096,0.03288622366057502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,3584,0.021644444929228887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,3584,0.03286400106218126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,3072,0.016200888488027785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,5120,0.025011556016074285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,3072,0.020608888732062448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,3072,0.030436442957984075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,2560,0.014011555247836642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,3584,0.018343110879262287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,2560,0.03064799971050686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,2048,0.011969777444998423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,2560,0.018588445252842374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,1536,0.009749333063761393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,2048,0.028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,2048,0.017239999439981248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,1024,0.008018666671382057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,1536,0.027088888817363318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,1536,0.01566844516330295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,1024,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,768,0.007053333024183909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,1024,0.022999111149046157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,768,0.012171555724408893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,768,0.02202044427394867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,512,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,512,0.020983111527231004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,256,0.0038773334688610504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,256,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,512,0.004879999905824661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,256,0.01983022193113963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,128,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,12288,128,0.01752711170249515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,64,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,128,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,12288,32,0.003528000166018804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,64,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,12288,32,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,65536,0.2164835532506307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,65536,0.15879021750556097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,51200,0.1731466717190213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,51200,0.12416799863179524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,65536,0.1593031088511149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,16384,0.06365333663092719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,16384,0.047909332646263965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,51200,0.12838489479488796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,16384,0.05887999799516466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,12288,0.04634044567743937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,12288,0.03974666529231601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,10240,0.04035200013054742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,12288,0.05021511183844673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,10240,0.03517511155870225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,8192,0.029306666718588933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,10240,0.04586844311820137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,8192,0.030639999442630347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,7168,0.026025777061780293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,8192,0.0407591097884708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,7168,0.026943110757403906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,6144,0.022896000080638464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,6144,0.024446222517225478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,6144,0.03604266709751553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,5120,0.019820445113711886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,5120,0.02240000002914005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,5120,0.034634666310416326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,4096,0.016413332687483895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,7168,0.03924355573124356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,3584,0.014740443891949125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,4096,0.03266400098800659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,4096,0.020638222495714825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,3072,0.01312888910373052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,3584,0.031018667750888403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,3584,0.019948444432682462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,2560,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,3072,0.018630221486091614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,3072,0.03045155604680379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,2048,0.010741333166758219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,2560,0.017304889029926725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,2560,0.02950577934583028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,1536,0.009106666677527958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,2048,0.01688088807794783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,2048,0.028296000427669946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,1024,0.007111999723646376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,1536,0.014863111906581454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,1536,0.026537777649031744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,768,0.006616000086069107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,1024,0.01275288893116845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,1024,0.022416000564893086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,512,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,768,0.012034667034943899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,768,0.02197155521975623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,256,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,512,0.020230222079488967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,512,0.012175111307038201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,128,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,256,0.019253333409627277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,256,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,64,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,128,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,10240,32,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,10240,128,0.01755466726091173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,64,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,10240,32,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,65536,0.16525067223442927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,51200,0.1315431065029568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,65536,0.1096773346265157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,51200,0.08603288729985555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,16384,0.053929779264662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,16384,0.03614844547377692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,65536,0.15299555990431044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,51200,0.12502933873070612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,12288,0.03279111120435927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,16384,0.05636355612013075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,10240,0.0284186667866177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,12288,0.030377778742048476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,12288,0.04868177904023064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,8192,0.023369777533743117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,10240,0.027222222752041284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,10240,0.044347554445266724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,8192,0.02296000056796604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,8192,0.04003644320699904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,7168,0.021054221524132624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,7168,0.02155911094612545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,7168,0.038027554750442505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,6144,0.0185208883550432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,6144,0.03537066777547201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,6144,0.020239111449983384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,5120,0.015831111205948722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,5120,0.01867377758026123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,5120,0.034056001239352755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,4096,0.013146667016877068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,4096,0.017487999465730455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,4096,0.03208622336387634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,3584,0.012319111161761813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,3584,0.017218665944205392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,3584,0.03056000007523431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,3072,0.011241777903503843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,3072,0.016228444046444364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,3072,0.030673778719372217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,2560,0.010085333552625444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,2560,0.015635554989178974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,2048,0.009204444785912832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,2048,0.014896000425020853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,2048,0.02852533260981242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,1536,0.007260444263617198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,1536,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,1536,0.024826667375034753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,1024,0.005564444594913059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,2560,0.02847733431392246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,1024,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,768,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,1024,0.02162666618824005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,768,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,768,0.020990222692489624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,512,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,512,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,512,0.020982222424613103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,256,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,256,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,256,0.019253333409627277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,128,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,128,0.01073333372672399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,8192,128,0.018592000007629395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,64,0.0029519999192820657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,8192,32,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,64,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,8192,32,0.01073066641887029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,65536,0.15523378054300943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,65536,0.10089688830905491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,51200,0.11596443918016221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,51200,0.07900533411237928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,65536,0.1518124474419488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,16384,0.043288889858457774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,16384,0.034019556310441755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,51200,0.12424888875749375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,12288,0.03302755620744493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,12288,0.028622223271263972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,16384,0.0558097759882609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,10240,0.028640889459186133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,10240,0.02571822206179301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,10240,0.0440977778699663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,8192,0.024138665861553613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,8192,0.03992088966899448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,12288,0.04868977930810717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,8192,0.021708443760871887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,7168,0.019016888406541612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,6144,0.0170133329100079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,7168,0.03723377651638455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,6144,0.019239111079110038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,5120,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,6144,0.03525244527392917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,7168,0.02093688812520769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,5120,0.017979555659823947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,5120,0.0336097776889801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,4096,0.01259200026591619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,3584,0.011632888681358762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,4096,0.03193333413865831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,3584,0.030252443419562444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,3072,0.009923555784755284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,3072,0.016168889072206285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,3584,0.01741422216097514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,2560,0.009016888837019602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,3072,0.029482665989134047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,2560,0.028035554620954726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,2048,0.007992000215583378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,2560,0.015328889091809591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,2048,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,4096,0.016896888613700867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,1536,0.007648000286685095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,1536,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,1024,0.004927999857399198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,2048,0.02832977639304267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,1024,0.01239999963177575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,768,0.0041991112132867174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,768,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,1024,0.023631999890009563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,512,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,768,0.021695110532972548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,512,0.020938666330443487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,512,0.011887110769748688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,1536,0.025628444221284654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,256,0.003305777700410949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,256,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,128,0.003133333391613431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,256,0.02096533278624217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,128,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,64,0.0028106667515304354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,7168,128,0.01943111088540819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,64,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,7168,32,0.003036444385846456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,7168,32,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,65536,0.09219021929634942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,65536,0.15108444955613878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,65536,0.12071911493937175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,51200,0.09609422418806289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,51200,0.07198666863971286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,16384,0.03721066647105747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,51200,0.12325778272416855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,16384,0.03257511059443156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,12288,0.028710222906536523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,16384,0.05553422371546427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,12288,0.047381334834628634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,12288,0.026896889011065166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,10240,0.02309866746266683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,10240,0.044545776314205594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,8192,0.022026666336589392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,8192,0.020991999242040846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,10240,0.02538577715555827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,7168,0.01963200006220076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,8192,0.039206223355399236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,7168,0.01976444489426083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,7168,0.03809244434038798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,6144,0.01752711170249515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,6144,0.03503200080659654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,6144,0.018948444061809115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,5120,0.015576000014940897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,5120,0.017904000149832833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,5120,0.03359199894799127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,4096,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,4096,0.016207110550668504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,3584,0.010773333410422007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,4096,0.03156000044610765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,3584,0.016139555308553908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,3584,0.03158488869667053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,3072,0.009505777723259395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,3072,0.029278222057554457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,3072,0.015676443775494892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,2560,0.008973333570692275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,2048,0.007795555724038019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,2560,0.014165333575672574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,2560,0.02922044528855218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,1536,0.0069777775141927935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,2048,0.02514666650030348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,2048,0.013378666506873237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,1024,0.0047404443224271136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,1536,0.023697778582572937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,1536,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,768,0.004259555704063839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,1024,0.02132266594303979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,1024,0.012084444363911947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,512,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,768,0.012059555285506778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,512,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,256,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,256,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,256,0.019665777683258057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,128,0.0034471110751231513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,768,0.0209840006298489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,128,0.010384889112578498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,64,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,128,0.017242666747834947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,6144,512,0.02132533325089349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,6144,32,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,64,0.01074577785200543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,65536,0.10904088947508071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,6144,32,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,65536,0.08293600214852227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,51200,0.08660888671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,51200,0.06600089205635919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,16384,0.03545688920550876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,65536,0.1504488918516371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,51200,0.12372177177005345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,16384,0.030004445049497817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,16384,0.055325332615110606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,12288,0.026151999831199646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,10240,0.02330933345688714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,12288,0.047041777107450694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,10240,0.02267733381854163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,10240,0.04299199912283155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,8192,0.019965334071053397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,8192,0.01957244508796268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,7168,0.017872000734011333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,8192,0.039177778694364764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,7168,0.01872711049185859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,7168,0.037809779246648155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,6144,0.01608533329433865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,6144,0.03472177849875556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,6144,0.018096889058748882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,5120,0.014085332552591959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,12288,0.024265777733590867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,5120,0.017282666431532968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,5120,0.03298844562636481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,4096,0.015477332803938123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,4096,0.03078311019473606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,3584,0.009677333136399588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,3584,0.015933333171738517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,4096,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,3072,0.008728000024954477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,3584,0.030794666873084173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,3072,0.02903999884923299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,2560,0.008239111138714684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,2560,0.014160000615649752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,2560,0.028346667687098186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,2048,0.0069475554757648045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,2048,0.01313333296113544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,3072,0.01482222146458096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,1536,0.005980444451173146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,2048,0.025192888246642217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,1536,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,1024,0.004584889031118817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,1536,0.02380888826317257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,1024,0.012094222009181976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,1024,0.02218577762444814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,768,0.004252444538805219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,768,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,768,0.02237333357334137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,512,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,512,0.011681777735551199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,512,0.020724443925751578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,256,0.0032995556377702286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,128,0.0033208889265855155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,256,0.010842666857772403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,256,0.020942222740915086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,64,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,128,0.011187555889288584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,5120,128,0.01815022197034624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,5120,32,0.003079111170437601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,64,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,65536,0.09037511216269599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,5120,32,0.010599999792046018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,65536,0.07775644461313884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,51200,0.07136533657709758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,51200,0.06298222144444783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,16384,0.028959999481836956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,65536,0.14997955163319907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,51200,0.1227137777540419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,12288,0.02125866711139679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,16384,0.02667644454373254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,16384,0.05523555477460226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,10240,0.018609777092933655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,12288,0.022239110536045496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,12288,0.04703288939264086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,10240,0.020782222350438435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,8192,0.015541333291265698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,10240,0.04277955492337545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,8192,0.018686221705542672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,7168,0.01423911088042789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,8192,0.03934222128656175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,6144,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,7168,0.018179555733998615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,7168,0.037442667616738215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,5120,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,6144,0.017539555827776592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,6144,0.03486311104562547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,4096,0.010027555955780877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,5120,0.01646133263905843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,5120,0.033356444703208074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,3584,0.009404444032245213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,4096,0.015608888533380298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,4096,0.03047910994953579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,3584,0.028939555088678997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,3072,0.008667555948098501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,3584,0.014650666051440768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,2560,0.007718222008811102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,3072,0.013843555417325763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,3072,0.027912888262006972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,2048,0.006046222315894232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,2560,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,2560,0.026455110973782007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,1536,0.0053386667536364655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,2048,0.01290933373901579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,2048,0.025733333494928148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,1536,0.012470222181744046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,1024,0.0042275554604000514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,1536,0.023706666297382776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,768,0.0035955554081334006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,1024,0.012096000214417776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,1024,0.023211555348502264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,512,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,768,0.021408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,768,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,256,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,512,0.020662221643659804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,512,0.011510222322411008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,256,0.010710222025712332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,256,0.020350222786267597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,64,0.0026586666289303037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,128,0.011526222030321756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,4096,128,0.018030222919252183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,64,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,4096,32,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,65536,0.07445333401362102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,4096,32,0.011147555377748279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,65536,0.07466222180260552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,51200,0.05979910824033949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,65536,0.14976444509294298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,51200,0.0606915553410848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,16384,0.02531911101606157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,16384,0.02562933332390255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,51200,0.12249867121378581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,12288,0.018983999888102215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,16384,0.05426311161783007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,12288,0.021742223037613764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,10240,0.01665511065059238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,12288,0.046542220645480685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,10240,0.020271999968422782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,8192,0.014246222045686511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,8192,0.018523555662896898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,8192,0.039445334010654025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,7168,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,7168,0.017830222845077515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,7168,0.03689688775274489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,6144,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,10240,0.04283644424544441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,6144,0.01716444392999013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,5120,0.011563555234008364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,6144,0.03552088803715176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,5120,0.015607110328144498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,4096,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,4096,0.01460977726512485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,4096,0.03143822153409322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,3584,0.009029332962301042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,3584,0.014187556174066333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,3584,0.03016177813212077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,3072,0.008309333688682979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,5120,0.032912890116373696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,3072,0.01347733371787601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,2560,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,3072,0.026757333013746474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,2560,0.013080889152155982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,2048,0.005913777897755305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,2560,0.02646488944689433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,2048,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,1536,0.00490488898422983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,2048,0.024749333659807842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,1536,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,1536,0.02347555591000451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,1024,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,1024,0.022038223014937505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,1024,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,768,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,768,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,768,0.021665778425004747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,512,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,512,0.011381333072980246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,512,0.02064533366097344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,256,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,256,0.020247111717859905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,256,0.01071999967098236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,128,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3584,128,0.018954666124449838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,128,0.010565333068370819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3584,32,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,64,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,65536,0.06572800212436251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3584,32,0.010474666953086853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,65536,0.07157778077655368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,51200,0.05197422371970283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,65536,0.1497759951485528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,51200,0.058111998769972056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,16384,0.021373333202468023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,51200,0.12304000059763591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,16384,0.024548444483015273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,16384,0.05442755421002706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,12288,0.021192888418833416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,10240,0.015271999769740634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,10240,0.01973777843846215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,10240,0.042803555727005005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,8192,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,8192,0.018208889497650992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,8192,0.039136889908048846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,7168,0.011953777737087674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,12288,0.04642666710747612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,7168,0.017498667041460674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,6144,0.011720889144473605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,7168,0.037371555964152016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,6144,0.01648977730009291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,5120,0.010648888846238455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,6144,0.03519644339879354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,5120,0.015059555570284525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,5120,0.03199022346072727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,4096,0.008403555386596257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,4096,0.014027555783589682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,4096,0.03133066826396518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,3584,0.008270222279760573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,12288,0.017248888810475666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,3072,0.007298666569921706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,3584,0.013779555757840475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,3584,0.0281368891398112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,3072,0.013155555559529198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,3072,0.026776888304286536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,2560,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,2560,0.01276711126168569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,2048,0.005567111074924469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,2048,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,2048,0.02482044365670946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,1536,0.00453955555955569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,1536,0.012076444096035428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,1536,0.023319111929999456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,1024,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,2560,0.026149332523345947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,1024,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,768,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,768,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,768,0.021324444148275588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,512,0.0034479999707804788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,512,0.011040888726711273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,512,0.020657777786254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,256,0.0031600000543726813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,1024,0.02200800014866723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,128,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,256,0.01070577816830741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,128,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,128,0.01792977750301361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,64,0.0028213332924577924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,3072,32,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,3072,256,0.020865778128306072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,65536,0.05703022082646688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,64,0.010790222220950656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,3072,32,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,51200,0.0462026662296719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,51200,0.05566844675276014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,65536,0.14955466323428684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,65536,0.06894399722417195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,16384,0.02029511166943444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,51200,0.12226222621070014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,16384,0.053878221246931285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,12288,0.015249777171346875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,12288,0.04671466681692335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,10240,0.014191110928853353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,12288,0.020742222666740417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,10240,0.01944977707333035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,8192,0.012071111136012606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,10240,0.042509333954917065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,8192,0.017851556340853374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,8192,0.038155555725097656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,16384,0.02408444384733836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,7168,0.017278222574128043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,6144,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,7168,0.03773333297835456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,6144,0.035671111610200666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,6144,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,5120,0.00980977796845966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,5120,0.033125334315829806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,5120,0.014607111612955729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,4096,0.008357333640257517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,7168,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,3584,0.007696000238259633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,4096,0.013493333425786761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,4096,0.02920088834232754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,3584,0.013471999930010902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,3072,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,3584,0.028215110301971436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,2560,0.00637511122557852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,3072,0.013210666676362356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,3072,0.027056889401541814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,2048,0.00534400012758043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,2560,0.02608266638384925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,2560,0.012795555922720166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,1536,0.004599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,2048,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,2048,0.02573066618707445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,1024,0.003962666624122196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,1536,0.011889778077602386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,1536,0.0243493335114585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,1024,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,768,0.0036071112586392295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,1024,0.022407111194398668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,512,0.0033928888539473214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,768,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,768,0.02269511090384589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,256,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,512,0.020669332808918424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,512,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,128,0.002944888960984018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,256,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,256,0.020449777444203694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,64,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2560,128,0.01923644377125634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2560,32,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,64,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,128,0.010419555836253695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,65536,0.04632888899909126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2560,32,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,51200,0.03828622235192193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,65536,0.14971644348568386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,51200,0.05382933219273885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,51200,0.1223271158006456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,16384,0.015104000767072042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,16384,0.023771555887328252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,16384,0.054049776660071484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,65536,0.06694577799903022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,12288,0.012703110774358114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,12288,0.046662221352259316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,12288,0.020385776956876118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,10240,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,10240,0.042789333396487765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,10240,0.01901244454913669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,8192,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,8192,0.03895733422703213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,8192,0.017235555582576327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,7168,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,7168,0.03593422306908502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,6144,0.010775999890433418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,7168,0.016021332807011075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,6144,0.015618667006492615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,5120,0.01018488903840383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,6144,0.03520266546143426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,5120,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,4096,0.00904622260067198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,5120,0.030618667602539062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,4096,0.013584888643688627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,3584,0.007119999991522894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,4096,0.028507557180192735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,3584,0.013530666629473368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,3072,0.006641777853171031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,3072,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,3584,0.028001778655582007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,2560,0.006011555592219035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,3072,0.027835554546780054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,2560,0.025785777303907607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,2560,0.012554667062229581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,2048,0.0053484443989064955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,2048,0.025739555557568867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,1536,0.004252444538805219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,2048,0.012243555651770698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,1536,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,1024,0.003964444415436851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,1536,0.02343822187847561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,1024,0.01187111106183794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,1024,0.022310222188631695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,768,0.0114720000161065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,768,0.02145333256986406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,512,0.0033617777129014335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,512,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,768,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,256,0.00300177786913183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,512,0.020664888951513503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,256,0.011319110790888468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,128,0.0029253332565228143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,256,0.019592000378502738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,128,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,64,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,2048,128,0.017990221579869587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,64,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,2048,32,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,65536,0.03649955656793382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,2048,32,0.011071110765139261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,65536,0.06409155660205416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,51200,0.029393778906928167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,65536,0.14958843919965956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,16384,0.01643200053109063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,51200,0.052581333451800875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,51200,0.12252977159288193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,12288,0.013341333303186627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,16384,0.022986667023764715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,16384,0.053632001082102455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,12288,0.02017688916789161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,10240,0.012560000022252401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,10240,0.01869777838389079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,8192,0.010674667027261523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,10240,0.04153777824507819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,8192,0.016599999533759225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,7168,0.010293333066834344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,8192,0.03673155440224542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,7168,0.01550044450494978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,6144,0.009451555709044138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,7168,0.035981333918041654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,12288,0.04590577880541483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,6144,0.014920888675583733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,5120,0.00902133352226681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,5120,0.014143110977278816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,6144,0.03366400135887994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,5120,0.030806223551432293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,4096,0.007690666450394525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,3584,0.007010666860474481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,4096,0.028786665863460962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,3584,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,4096,0.013463999662134381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,3072,0.006319111006127463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,3584,0.027437332603666518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,2560,0.005886222339338726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,3072,0.027454222242037456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,3072,0.012850667039553324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,2048,0.005361777626805835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,2560,0.02624622152911292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,2048,0.025409777959187824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,2048,0.01255733354224099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,1536,0.004640000147951974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,1536,0.011945777469211154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,1536,0.023430221610599097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,2560,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,1024,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,1024,0.011622221933470832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,768,0.0036515556275844574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,1024,0.022271111607551575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,768,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,512,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,768,0.021355556117163763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,512,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,256,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,512,0.02032355633046892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,256,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,256,0.019690665933820937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,128,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1536,128,0.017987555927700467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,64,0.010089777410030365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1536,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,65536,0.02557777696185642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1536,32,0.010398222340477837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,65536,0.06140889061821831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,65536,0.14963555335998535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,51200,0.021776888105604384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,51200,0.04994577831692166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,16384,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,51200,0.12201333045959473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,16384,0.02238666680124071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,12288,0.010637332995732626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,12288,0.01889155473974016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,16384,0.05256711112128364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,10240,0.009725333087974125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,12288,0.044919110006756253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,10240,0.04152266515625848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,10240,0.01776355504989624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,8192,0.016016888949606154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,8192,0.036208887894948326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,7168,0.008114666574531132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,8192,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,7168,0.015638222297032673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,6144,0.007631999750932057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,7168,0.03393155667516921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,6144,0.0321395562754737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,5120,0.007057777709431118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,6144,0.0147724449634552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,5120,0.014166222678290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,4096,0.00720888872941335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,5120,0.029992000924216375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,4096,0.013518222504191928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,3584,0.006888000087605582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,4096,0.02803200152185228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,3584,0.013165333204799227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,3072,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,3584,0.027896000279320612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,3072,0.0127368892232577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,2560,0.005649777750174205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,3072,0.026200888885392085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,2560,0.01253600004646513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,2048,0.005910222315125995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,2560,0.026147555973794725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,2048,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,1536,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,1536,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,2048,0.02446222139729394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,1024,0.00434933313065105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,1536,0.02342044479317135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,1024,0.021021333005693223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,768,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,1024,0.011604444848166572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,768,0.011123555401961008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,512,0.0035662220584021676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,512,0.011188444164064197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,256,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,256,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,512,0.021048888564109802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,128,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,256,0.01886577738655938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,128,0.017266665895779926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,128,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,64,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,1024,32,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,1024,768,0.021132444341977436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,65536,0.02720533311367035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,32,0.010415111151006486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,1024,64,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,51200,0.021341333786646526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,65536,0.14972355630662706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,65536,0.060602664947509766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,51200,0.04916266600290934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,16384,0.011144888897736868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,51200,0.12224711312188043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,12288,0.009377777576446533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,16384,0.022054221895005967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,16384,0.052490668164359204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,12288,0.018533334136009216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,10240,0.0087013335691558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,12288,0.04532000091340807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,10240,0.01719199948840671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,8192,0.008584000170230865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,10240,0.04116977916823493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,8192,0.015927111109097798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,7168,0.008028444316652086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,8192,0.036954667833116316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,7168,0.015174221661355762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,7168,0.03536444571283128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,6144,0.007630222373538547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,6144,0.031756444109810725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,6144,0.0151573336786694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,5120,0.0069475554757648045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,5120,0.030574222405751545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,4096,0.007335999773608313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,5120,0.014511111709806653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,4096,0.013182222015327878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,3584,0.0069324444565508105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,4096,0.028499556912316218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,3584,0.013077333569526672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,3584,0.028838220569822524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,3072,0.00665511108107037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,3072,0.026537777649031744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,2560,0.006265777680608962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,3072,0.013253333667914072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,2560,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,2048,0.004969777746333016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,2560,0.02716444432735443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,2048,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,1536,0.004527111020353106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,2048,0.025440888272391424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,1536,0.0118231111102634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,1536,0.024338665935728285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,1024,0.004048000193304486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,1024,0.021317332983016968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,1024,0.0120000003112687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,768,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,768,0.020301333732075162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,768,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,512,0.0035004446076022256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,512,0.0196079992585712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,512,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,256,0.0032053333189752367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,256,0.01889244384235806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,256,0.011329777538776398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,128,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,768,128,0.017504000001483493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,128,0.011356444822417365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,64,0.0028231110837724474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,768,32,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,64,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,768,32,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,65536,0.05930755535761515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,65536,0.1492533286412557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,51200,0.0159599996275372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,51200,0.04862222075462341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,51200,0.1220888826582167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,65536,0.018574222922325134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,16384,0.009735110733244155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,12288,0.008968888885445064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,16384,0.05237422386805216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,16384,0.02148977749877506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,12288,0.04384444488419426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,10240,0.008084444536103142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,10240,0.01692800058258904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,8192,0.007670222057236566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,8192,0.015778667396969266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,10240,0.040877332290013633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,7168,0.007734222544564142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,8192,0.035980443159739174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,7168,0.015632889337009855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,7168,0.0336168905099233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,6144,0.00740444411834081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,6144,0.014465777410401238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,12288,0.018612444400787354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,5120,0.006912888752089606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,6144,0.03329599897066752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,4096,0.007387555307812161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,5120,0.03050044510099623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,5120,0.014155555102560254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,3584,0.0069475554757648045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,4096,0.028740445772806805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,4096,0.013749333719412485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,3072,0.006622222148709827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,3584,0.02706844442420536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,3072,0.013100444442696042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,2560,0.006599111275540457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,3072,0.026452443665928308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,2560,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,2048,0.006285333385070165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,2048,0.012631999949614206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,2048,0.025040000677108765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,1536,0.004540444248252445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,3584,0.013402666482660504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,1536,0.012126222252845764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,1024,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,1536,0.024076445235146418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,1024,0.022338666849666174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,768,0.003504000190231535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,2560,0.025077333052953083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,1024,0.011975111232863532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,768,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,512,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,768,0.022685334086418152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,512,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,512,0.02158311174975501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,256,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,256,0.020986666282018025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,512,128,0.01891288823551602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,128,0.011384888655609555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,64,0.0025457777082920074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,512,32,0.0025724443710512584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,64,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,512,32,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,65536,0.014670222997665405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,65536,0.058592001597086586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,51200,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,65536,0.1493475569619073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,51200,0.04641155401865641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,51200,0.1219751172595554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,16384,0.008664888640244802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,16384,0.05172622203826904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,16384,0.021287999219364587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,12288,0.008005333443482717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,12288,0.0188711103465822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,12288,0.043903999858432345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,10240,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,10240,0.017312889297803242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,10240,0.040720889965693154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,8192,0.0069422221018208405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,8192,0.03637866510285272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,8192,0.01620533400111728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,7168,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,7168,0.01575288838810391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,7168,0.03439555565516154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,6144,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,6144,0.014781332678265043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,5120,0.005927111125654644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,6144,0.03330577744377984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,5120,0.014124444789356656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,5120,0.030930668115615845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,4096,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,4096,0.013825777504179211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,4096,0.029190222422281902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,3584,0.006304889089531369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,3584,0.013100444442696042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,3584,0.028877334462271795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,3072,0.006063111126422882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,3072,0.0277839998404185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,3072,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,2560,0.006593777901596493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,2560,0.02630222174856398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,2048,0.0058204444746176405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,2048,0.012466666599114736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,2048,0.025765332910749648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,1536,0.0041911109454101985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,1536,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,1536,0.02418488926357693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,1024,0.0038542221817705366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,2560,0.012776888906955719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,1024,0.01181777815024058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,768,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,1024,0.02306488818592495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,768,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,512,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,512,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,512,0.020678222179412842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,256,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,256,0.011124444504578909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,128,0.0026400000270869997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,128,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,256,0.019925332731670804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,768,0.022652443912294176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,64,0.00272444449365139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,256,128,0.018911111685964797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,256,32,0.002532444480392668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,65536,0.014278221461508008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,32,0.011156444748242697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,256,64,0.01147822207874722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,51200,0.012472000386979846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,65536,0.14782044622633192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,51200,0.045888887511359326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,65536,0.05500266618198819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,16384,0.008089777496125963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,51200,0.11980088551839192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,12288,0.007640889121426477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,16384,0.02083288960986667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,16384,0.05178933342297872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,12288,0.018631110588709515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,10240,0.007081777685218387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,12288,0.04384444488419426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,8192,0.006695111178689533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,10240,0.017332444588343304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,10240,0.04088088870048523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,7168,0.006583111153708563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,8192,0.0366542206870185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,8192,0.01588888963063558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,6144,0.00600355573826366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,7168,0.03417066733042399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,7168,0.015815110670195687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,6144,0.014815110299322339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,5120,0.006004444426960415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,6144,0.03333955672052171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,5120,0.014160000615649752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,5120,0.03087377879354689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,4096,0.005913777897755305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,4096,0.0288426677385966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,3584,0.006278222219811545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,4096,0.013754666679435305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,3584,0.013360888593726687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,3584,0.028873778051800195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,3072,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,3072,0.027849776877297297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,3072,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,2560,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,2560,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,2560,0.02607911162906223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,2048,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,2048,0.012728888955381183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,2048,0.025450666745503742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,1536,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,1536,0.02439555525779724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,1024,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,1024,0.012048888537618848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,1024,0.023330666952663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,768,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,1536,0.004271111140648524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,768,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,512,0.0031644445326593188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,768,0.021355556117163763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,512,0.021029333273569744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,512,0.011657777759763928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,256,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,256,0.011376889215575324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,128,0.0029315555261241067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,128,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,128,0.018964444597562153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,64,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,64,0.011380444798204633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,128,32,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,32,128,256,0.020269332660569083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,65536,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,128,32,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,51200,0.010008888940016428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,65536,0.054119110107421875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,16384,0.0070053330726093715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,51200,0.04460710949367947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,12288,0.006948444578382704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,16384,0.020991999242040846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,10240,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,12288,0.0188719994492001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,8192,0.006606222026877933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,10240,0.017264000243610807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,7168,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,7168,0.015499555402331881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,6144,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,8192,0.01628800067636702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,5120,0.006256888724035687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,6144,0.014841778410805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,4096,0.005961777849329843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,5120,0.014554666148291694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,3584,0.006064888917737537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,4096,0.013653332988421122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,3072,0.005910222315125995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,3584,0.013535999589496188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,2560,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,3072,0.013106666505336761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,2048,0.005605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,2048,0.012497777740160624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,1536,0.004239999999602635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,1024,0.003936888857020272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,1536,0.012061333490742577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,768,0.0032799999333090256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,1024,0.011360888679822287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,512,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,768,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,256,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,512,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,128,0.00282222218811512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,256,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,64,0.002507555608948072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,128,0.01051999959680769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,64,32,0.002552888873550627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,64,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,65536,0.008960888617568547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,2560,0.012979555461141797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,64,32,0.010024000373151567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,51200,0.008714666797055138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,65536,0.054082665178510875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,16384,0.006679111056857639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,51200,0.04387911160786947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,12288,0.006042666733264923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,16384,0.020991999242040846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,10240,0.006457777900828256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,12288,0.017880888448821176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,8192,0.005776888794369168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,10240,0.017200888858901132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,7168,0.006584888945023219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,8192,0.01553600033124288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,6144,0.005655111124118169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,7168,0.01554666625128852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,5120,0.005913777897755305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,6144,0.01442311041884952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,4096,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,5120,0.014103111293580798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,3584,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,4096,0.013400889105266996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,3072,0.00555288874440723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,3584,0.013275555438465543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,2560,0.006249777972698212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,3072,0.012713777522246042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,2048,0.005583999885453119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,2560,0.012727110750145383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,1536,0.004368000146415499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,2048,0.012085333466529846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,1024,0.003896000070704354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,1536,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,768,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,1024,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,512,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,768,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,256,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,512,0.010727110836240979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,128,0.002584888910253843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,256,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,64,0.002469333302643564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,128,0.010487111078368293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,32,32,32,0.002513777878549364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,64,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,32,32,32,0.010077333284748925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,51200,0.49720356199476456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,51200,1.0623849232991536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,16384,0.3649368815951877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,16384,0.16853066285451254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,12288,0.25998134083218044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,10240,0.21760443846384683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,12288,0.13046222262912327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,10240,0.11213421821594238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,8192,0.1795626613828871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,7168,0.15377867221832275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,8192,0.09082221984863281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,7168,0.12093688382042779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,6144,0.13720088534884983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,5120,0.1100533339712355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,6144,0.09458933273951213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,5120,0.07396355602476332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,4096,0.08986666467454697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,3584,0.07905244164996676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,4096,0.07193599806891547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,3072,0.06780266761779785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,3584,0.06039555867513021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,3072,0.05273155702484978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,2560,0.04744889007674324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,2048,0.04745777779155307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,2048,0.03815199931462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,1536,0.0361768901348114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,1536,0.03210577699873183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,1024,0.02562933332390255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,2560,0.05716977516810099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,768,0.020192000601026747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,1024,0.02535733415020837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,512,0.014104000396198697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,768,0.02232355541653103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,512,0.019107555349667866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,256,0.009293333523803288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,128,0.005991999887757831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,64,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,256,0.015630222029156156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,65536,32,0.004622222234805425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,64,0.014092443717850579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,128,0.013094222380055321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,65536,32,0.014665777484575907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,65536,0.5260640250311958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,51200,0.7976329061720105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,65536,1.0425146950615776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,16384,0.27183911535474986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,51200,0.40285245577494305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,16384,0.13770843876732722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,12288,0.1073955562379625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,12288,0.1993457741207547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,10240,0.16691466172536215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,8192,0.13465155495537653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,8192,0.075527999136183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,7168,0.1173279947704739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,6144,0.1023448904355367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,10240,0.09275822507010566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,7168,0.08164888620376587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,5120,0.08710933393902248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,5120,0.06241155332989163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,4096,0.0733848876423306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,6144,0.07646755377451579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,3584,0.06303289201524523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,4096,0.06198755237791273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,3072,0.054077333874172635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,3584,0.05144355694452921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,2560,0.046550220913357206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,3072,0.04623644550641378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,2048,0.03829688827196757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,2560,0.03977955712212457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,1536,0.02901866700914171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,2048,0.033556444777382746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,1024,0.021631111701329548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,1536,0.02773333258099026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,768,0.016215110818545025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,1024,0.021925333473417494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,512,0.012095111111799875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,768,0.02012711101108127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,256,0.007648000286685095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,512,0.017606221967273288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,128,0.0042160000238153665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,256,0.014504889647165934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,64,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,128,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,51200,32,0.003922666526503033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,64,0.014084445105658637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,51200,32,0.013417777915795645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,65536,0.32903822263081867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,65536,0.18332355552249482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,51200,0.25493065516153973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,16384,0.08857866790559556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,51200,0.1413066652086046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,12288,0.06218666500515408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,16384,0.055312001042895846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,12288,0.04483022292455038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,10240,0.052742222944895424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,8192,0.042805333932240806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,10240,0.03996710975964864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,7168,0.037901331981023155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,8192,0.034276445706685386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,6144,0.033391998873816595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,7168,0.03868444429503547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,5120,0.028688000308142766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,6144,0.03590844406021966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,4096,0.02366222275627984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,5120,0.030584000878863867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,3584,0.02106488909986284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,3072,0.018624888526068795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,4096,0.027482666903071936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,2560,0.016042666302786935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,3072,0.023390221926901076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,2048,0.013395555317401886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,2560,0.02202933364444309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,1536,0.010668444136778513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,2048,0.01921333372592926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,3584,0.026686223016844854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,1536,0.017122666041056316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,768,0.007101332975758447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,768,0.013914666242069669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,512,0.00509066672788726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,1024,0.015669332610236276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,512,0.012763555679056378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,256,0.012191111014948951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,128,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,128,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,64,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,1024,0.00852177788813909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,32,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,64,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,16384,32,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,65536,0.24531822734408906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,16384,256,0.003735110991530948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,51200,0.19190311431884766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,16384,0.07262933254241943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,51200,0.12278489271799724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,16384,0.04907200071546766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,12288,0.05366488960054186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,10240,0.04636889033847385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,12288,0.04004977809058295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,8192,0.03934311204486423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,65536,0.15647110674116346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,7168,0.034859554635153875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,8192,0.03120622369978163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,10240,0.03608622153600057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,6144,0.030673778719372217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,5120,0.024077332682079736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,7168,0.03330755564901564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,4096,0.0199955552816391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,5120,0.02753777801990509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,6144,0.03252622154023912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,3584,0.01757066614098019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,3072,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,4096,0.025842666625976562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,2560,0.013739555246300168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,3072,0.02234577801492479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,2048,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,2560,0.020404444800482858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,3584,0.02416977783044179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,1536,0.009881777895821465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,1024,0.008039111064540016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,2048,0.01824711097611321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,768,0.006691555596060223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,1024,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,512,0.004512000001139111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,1536,0.016843555702103507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,768,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,256,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,512,0.012485333614879184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,128,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,256,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,64,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,128,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,12288,32,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,64,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,12288,32,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,65536,0.2086684438917372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,65536,0.15137244595421687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,51200,0.1639688942167494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,16384,0.06378755304548475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,51200,0.1179315514034695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,12288,0.047797332207361855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,16384,0.04740177922778659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,10240,0.04088711076312595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,12288,0.038740442858801946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,8192,0.035070220629374184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,10240,0.034874667723973594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,7168,0.03126577867401971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,8192,0.030143999391131934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,6144,0.022794667217466567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,7168,0.031354665756225586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,5120,0.020055999358495075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,6144,0.02716711163520813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,4096,0.0162773331006368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,5120,0.02436622149414486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,3584,0.014969777729776172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,4096,0.022659555077552795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,3072,0.013024888932704926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,3584,0.0214711113108529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,2560,0.011609777808189392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,3072,0.019573332534896005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,2048,0.009879999690585667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,2560,0.01873333255449931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,2048,0.01629244453377194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,1536,0.008570666942331526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,1024,0.007416889071464539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,1536,0.01495822270711263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,768,0.005568000177542369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,768,0.012690666649076672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,512,0.004569777597983678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,512,0.012315555579132505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,256,0.0036480000449551474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,256,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,128,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,128,0.010673777924643623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,64,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,1024,0.013457777599493662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,10240,32,0.0031724443866146933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,64,0.010845333337783813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,10240,32,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,65536,0.1703182193968031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,65536,0.10176889101664226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,51200,0.13508000638749865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,16384,0.05259911219278971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,51200,0.08006755510965984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,12288,0.032460444503360324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,16384,0.03513333201408386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,10240,0.028334223561816748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,12288,0.029789331886503432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,8192,0.02314400010638767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,10240,0.026820444398456152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,7168,0.020985777179400127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,8192,0.02230577833122677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,6144,0.018633777896563213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,7168,0.02420799930890401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,5120,0.016356445021099515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,6144,0.022355554832352534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,4096,0.013213333156373767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,5120,0.02017333275742001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,3584,0.01220355596807268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,4096,0.01928711103068458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,3072,0.010926222635640038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,3584,0.018331555856598746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,2560,0.009852444132169088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,3072,0.01761244402991401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,2560,0.01682044400109185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,2048,0.014838222000333997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,1536,0.007111111448870764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,1536,0.013448889056841532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,1024,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,1024,0.012440889245933957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,768,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,768,0.012387555506494312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,512,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,2048,0.008525333470768398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,256,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,512,0.012040889097584618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,128,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,256,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,64,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,128,0.01036177741156684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,8192,32,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,64,0.011345778074529437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,8192,32,0.010737777584128909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,65536,0.15608266989390054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,51200,0.11521777841779922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,65536,0.09328266647126938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,16384,0.04334133201175266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,16384,0.033166223102145724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,51200,0.07327377796173096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,12288,0.03345600101682875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,12288,0.027824888626734417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,10240,0.029511110650168523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,10240,0.024570667081409033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,8192,0.024467556013001338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,8192,0.021239999267790053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,7168,0.02199288871553209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,7168,0.023361777265866596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,6144,0.016902221573723685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,6144,0.02288355595535702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,5120,0.014911111858155994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,5120,0.01961688862906562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,4096,0.012221333053376941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,4096,0.019282667173279654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,3584,0.011211555865075854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,3584,0.01792533364560869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,3072,0.01014311114947001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,3072,0.016862221890025668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,2560,0.00906133320596483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,2560,0.016188444362746347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,2048,0.008170666793982187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,2048,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,1536,0.006981333096822103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,1536,0.013202666408485837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,1024,0.004628444297446145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,1024,0.012478222449620565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,768,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,768,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,512,0.003567999849716822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,512,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,256,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,256,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,128,0.0033964444365766314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,64,0.003097777772280905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,128,0.010424888796276517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,7168,32,0.003165333428316646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,32,0.010408889088365767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,7168,64,0.011221333510345884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,65536,0.12446044551001655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,51200,0.09866666793823242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,65536,0.08545777532789443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,51200,0.0668684442838033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,16384,0.03787822193569607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,12288,0.02907022171550327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,16384,0.031074666314654883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,10240,0.02591466738118066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,10240,0.02269777821169959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,8192,0.02109066645304362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,12288,0.02666933337847392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,7168,0.019504888190163504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,8192,0.020270221763186984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,6144,0.017232000827789307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,6144,0.02201777696609497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,7168,0.023902222514152527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,5120,0.01555999947918786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,4096,0.011622221933470832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,5120,0.0192239996459749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,4096,0.01826133330663045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,3584,0.010773333410422007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,3072,0.009749333063761393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,3584,0.01756977703836229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,2560,0.008810666700204214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,2560,0.015432000160217285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,3072,0.016794666647911072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,2048,0.007865777446164025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,1536,0.00637155564294921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,1536,0.013070222404268054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,1024,0.004365333252482944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,1024,0.01236533373594284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,768,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,768,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,512,0.003640888879696528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,512,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,256,0.003306666596068276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,256,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,2048,0.014128888646761576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,128,0.0032951111594835916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,64,0.002943111169669363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,128,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,64,0.01036977767944336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,6144,32,0.002944888960984018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,6144,32,0.010392888552612728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,65536,0.11097688807381524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,51200,0.088228444258372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,65536,0.07717866367763944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,16384,0.03479555580351088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,16384,0.028840889533360798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,51200,0.061943113803863525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,12288,0.026446221603287592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,10240,0.022482666704389784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,10240,0.02131555477778117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,8192,0.019718221492237516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,12288,0.023781332704755995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,7168,0.01776355504989624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,8192,0.01926577753490872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,6144,0.01575644479857551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,6144,0.020770667327774894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,7168,0.02273955610063341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,5120,0.01387288918097814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,4096,0.010407999985747868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,5120,0.01794666714138455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,4096,0.017941332525677152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,3584,0.009831111464235518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,3584,0.016960889101028442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,3072,0.00904799997806549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,3072,0.016021332807011075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,2560,0.008011555506123437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,2560,0.01497688889503479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,2048,0.007170666423108842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,2048,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,1536,0.00611111107799742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,1024,0.004571555389298333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,1536,0.01331555595000585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,768,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,1024,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,512,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,768,0.012477333347002665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,256,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,512,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,128,0.003165333428316646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,256,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,64,0.0029360000044107437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,128,0.010407110883129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,5120,32,0.002943111169669363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,64,0.011132444772455426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,65536,0.08975378010008071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,5120,32,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,51200,0.07360266976886325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,65536,0.07133244143591987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,16384,0.027892443868849013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,51200,0.05702577696906196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,12288,0.02111377815405528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,16384,0.026530666483773127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,10240,0.018783999813927543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,12288,0.02197511163022783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,8192,0.015695111619101632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,10240,0.020647999313142564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,8192,0.018439110782411363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,7168,0.021514667405022517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,6144,0.012759110993809171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,6144,0.020447999238967896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,5120,0.011242667006121742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,4096,0.009655111365848118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,5120,0.018438221679793466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,3584,0.008854221966531541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,7168,0.014051554931534661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,3072,0.008075555165608725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,4096,0.016229333149062265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,2560,0.007727111379305522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,3584,0.015519110692871941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,2048,0.005946666830115848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,3072,0.014482667048772177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,2560,0.013561777770519257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,1536,0.004634666774008009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,2048,0.013208889298968844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,1024,0.003941333128346337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,768,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,1536,0.012475555141766867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,512,0.003591999825504091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,1024,0.011518222590287527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,256,0.003173333282272021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,768,0.012412444584899478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,512,0.011749332977665795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,128,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,64,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,128,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,4096,32,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,256,0.010768888725174798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,64,0.011184888581434885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,4096,32,0.010760000182522668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,65536,0.07710222403208415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,51200,0.06005422274271647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,16384,0.02450399928622776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,65536,0.06765600045522054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,12288,0.019685332973798115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,51200,0.05500355694029066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,16384,0.024509333901935156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,10240,0.017699556218253244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,12288,0.02166755497455597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,8192,0.015085332923465304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,10240,0.019896888070636325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,8192,0.01828266680240631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,6144,0.012000888586044312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,7168,0.019889778561062284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,5120,0.010584000084135266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,6144,0.01891377733813392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,5120,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,4096,0.009095999929640029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,3584,0.008594666918118795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,4096,0.014856000741322836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,3072,0.00792266676823298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,3584,0.014886221951908536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,2560,0.0069120000633928515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,3072,0.013855111267831592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,2048,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,7168,0.013740444348918067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,2560,0.013999111122555204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,1536,0.004581333448489507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,1024,0.003938666648334927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,2048,0.013032000097963544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,1536,0.0129466669427024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,768,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,1024,0.011680000358157687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,512,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,768,0.012422222230169507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,256,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,512,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,256,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,64,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,128,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3584,32,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,64,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,65536,0.067193779680464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3584,32,0.011054221954610614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,51200,0.05074844426578946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,65536,0.06494933366775513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,16384,0.021799999806616042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,51200,0.052779555320739746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,16384,0.024079999989933435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,12288,0.016566221912701923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,10240,0.015455111861228943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,12288,0.020922667450375024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,8192,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,10240,0.019342222147517733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,7168,0.01216533366176817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,8192,0.01798311041461097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,6144,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,7168,0.01925244430700938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,5120,0.010441777606805166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,6144,0.01759111053413815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,4096,0.008729777402347988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,5120,0.01607644392384423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,3584,0.008040000167157915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,4096,0.014208000567224292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,3072,0.007742221984598372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,3584,0.013955555028385587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,2560,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,3072,0.013532444834709167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,2048,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,2560,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,1536,0.0052915554907586836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,2048,0.012744888663291931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,1024,0.004565333326657613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,1536,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,768,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,1024,0.011794666449228922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,512,0.003809777812825309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,768,0.011875555747085147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,512,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,256,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,128,0.002966222249799304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,256,0.011482666763994427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,64,0.0028311111446883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,128,0.010418666733635796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,3072,32,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,64,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,65536,0.05618844429651896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,3072,32,0.011133333047231039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,51200,0.045829332537121244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,65536,0.06248799959818522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,16384,0.020089776979552377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,51200,0.05091999967892965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,12288,0.015214223000738354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,16384,0.023552889625231426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,10240,0.01405422223938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,12288,0.02038666605949402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,8192,0.012269333004951477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,10240,0.0191920002301534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,7168,0.011505777637163797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,8192,0.017280000779363845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,6144,0.010575110713640848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,7168,0.01792977750301361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,5120,0.010056888891590966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,6144,0.016741333736313712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,4096,0.00851999968290329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,5120,0.014742222097184924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,3584,0.007778666913509369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,4096,0.013825777504179211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,3072,0.007649777664078607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,3584,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,2560,0.006607111129495833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,3072,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,2048,0.005655999812814925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,2560,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,1536,0.00519466648499171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,1024,0.004393777913517422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,2048,0.012422222230169507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,768,0.003956444561481476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,1536,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,512,0.003551111039188173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,1024,0.011542222566074796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,256,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,768,0.011759999725553723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,128,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,512,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,256,0.011014222270912595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2560,32,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,64,0.010422222316265106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,128,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2560,32,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,65536,0.06027555465698242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,51200,0.0388648874229855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,51200,0.04972177743911743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,16384,0.015215999550289579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,16384,0.023056000471115112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,65536,0.04727200004789564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,12288,0.012879110872745514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,10240,0.013724444641007317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,12288,0.02008888953261905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,10240,0.018947554959191214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,8192,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,7168,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,8192,0.01734044485621982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,6144,0.01070577816830741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,6144,0.014948444234000312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,5120,0.009695111049546136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,5120,0.014505777094099255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,4096,0.008103999826643202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,4096,0.01389244364367591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,3584,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,7168,0.016424889365832012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,3072,0.006211555666393704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,3584,0.013805333111021252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,2560,0.005684444473849402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,3072,0.012755555411179861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,2048,0.004945777770545748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,2560,0.013167111410035027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,1536,0.004631999880075455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,1536,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,1024,0.003930666794379552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,1024,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,768,0.0038933331767717996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,768,0.012148444851239523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,512,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,2048,0.01205599970287747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,256,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,512,0.011930666863918304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,128,0.0029360000044107437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,256,0.010791999598344168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,64,0.002567111204067866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,128,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,2048,32,0.0029377777957253983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,64,0.01038044442733129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,65536,0.036441776487562395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,2048,32,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,65536,0.05819288889567057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,51200,0.02946044339074029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,16384,0.015590222345458137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,51200,0.04764444298214383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,12288,0.014084445105658637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,16384,0.022758222288555566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,10240,0.01183644433816274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,12288,0.019952888290087383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,8192,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,10240,0.018242667118708294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,7168,0.010659555594126383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,8192,0.015882665912310284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,7168,0.015597333510716757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,6144,0.009758222434255812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,5120,0.009209777745935652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,6144,0.014901333385043673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,4096,0.008295111358165741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,5120,0.013807999591032663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,3584,0.007647111184067196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,4096,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,3584,0.013179555535316467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,3072,0.01278222186697854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,2560,0.00554044461912579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,2560,0.012506666282812754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,2048,0.0053324442770746015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,1536,0.004295999805132548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,2048,0.012311110893885294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,1024,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,1536,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,768,0.0036106668412685394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,3072,0.006628444625271692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,512,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,1024,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,256,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,768,0.01186222251918581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,512,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,256,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,64,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1536,32,0.002805333377586471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,128,0.01034755590889189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,64,0.010800888968838586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,65536,0.02609777781698439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,51200,0.02162133322821723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1536,32,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,16384,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,65536,0.05572622352176242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,12288,0.010943111446168689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,51200,0.0455991095966763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,16384,0.02204088866710663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,10240,0.010301333334710863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,12288,0.018647111124462552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,8192,0.00906666699382994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,10240,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,8192,0.01586844523747762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,7168,0.008295999632941352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,6144,0.009008888569143083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,7168,0.015478221906556023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,5120,0.0076657773719893555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,6144,0.014889778362380134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,4096,0.007316444483068254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,5120,0.014156444205178155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,3584,0.006663110935025745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,4096,0.013704000247849358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,3584,0.013506666653686099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,3072,0.006285333385070165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,2560,0.005599110904667113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,3072,0.01309511148267322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,2048,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,2560,0.013050666285885705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,1536,0.004921777794758479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,2048,0.012322666744391123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,1024,0.0042035554846127825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,1536,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,768,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,1024,0.01181244436237547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,512,0.00328177772462368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,768,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,512,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,256,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,128,0.011236444115638733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,64,0.0025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,64,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,32,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,1024,32,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,1024,256,0.00327822214199437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,65536,0.0548017786608802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,51200,0.019667555888493855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,51200,0.04477599925465054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,16384,0.01075644459989336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,16384,0.021339555581410725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,12288,0.010768888725174798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,65536,0.023922666907310486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,10240,0.009688888986905416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,12288,0.018248889181349013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,8192,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,10240,0.016895999511082966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,7168,0.0094035557574696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,8192,0.0158933334880405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,6144,0.008620444271299574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,7168,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,5120,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,6144,0.014853333433469137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,4096,0.007292444507280986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,5120,0.014112000664075216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,3584,0.006604444649484422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,4096,0.013550221920013428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,3072,0.006339555399285422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,3584,0.013432000246312885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,2560,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,3072,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,2048,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,2560,0.012746666868527731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,1536,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,2048,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,1024,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,1536,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,768,0.003646222253640493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,1024,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,512,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,768,0.012000888586044312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,256,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,512,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,256,0.01126933346192042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,64,0.002817777709828483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,128,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,768,32,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,64,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,768,32,0.011029333704047732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,65536,0.017623111605644226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,51200,0.01627377834584978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,16384,0.010010666317409938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,65536,0.05372177892261081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,51200,0.044171555174721606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,12288,0.010110222631030613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,16384,0.020584000481499564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,10240,0.008979555633332994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,8192,0.008019555773999956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,12288,0.017854221993022494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,7168,0.008619555168681675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,10240,0.01719911065366533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,6144,0.008029333419269985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,8192,0.01586666703224182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,5120,0.007600000335110559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,7168,0.01516355574131012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,4096,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,6144,0.014650666051440768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,3584,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,5120,0.014508444401952954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,4096,0.013080000049538083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,3072,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,3584,0.013474666410022311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,2560,0.00565422202150027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,3072,0.012706666356987424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,2048,0.004920000003443824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,2560,0.01277511070171992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,1536,0.0042088888585567474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,2048,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,1024,0.003863111138343811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,1536,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,768,0.0034871109657817413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,1024,0.011364444262451597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,512,0.0034702221552530923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,768,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,256,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,512,0.011013333168294696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,128,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,256,0.011209777659840055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,64,0.002514666567246119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,128,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,512,32,0.002570666579736604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,64,0.01070577816830741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,65536,0.016737777325842116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,512,32,0.010388444695207808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,51200,0.015495111544926962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,65536,0.05060355530844795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,16384,0.009348444640636444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,51200,0.0411102208826277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,12288,0.009406222237481011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,16384,0.020605333977275424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,10240,0.008795555267069075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,12288,0.017901332841979135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,8192,0.007985778152942657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,10240,0.01700799994998508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,7168,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,8192,0.015493333339691162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,6144,0.006950221955776215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,7168,0.015159999330838522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,5120,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,6144,0.014215999179416232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,4096,0.0062773335311147906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,5120,0.014109333356221518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,3584,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,4096,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,3072,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,3584,0.013424889081054263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,2560,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,3072,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,2048,0.005924444645643234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,2560,0.01273155543539259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,1536,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,2048,0.011546666423479715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,1024,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,1536,0.011957333319716983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,768,0.003519999898142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,1024,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,512,0.0033777778347333274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,768,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,256,0.003154666680428717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,512,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,128,0.0027955555253558685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,256,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,64,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,128,0.010057777994208867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,256,32,0.0027991111079851785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,64,0.010693333215183683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,65536,0.016181333197487723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,256,32,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,51200,0.014925332532988654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,65536,0.049392888943354286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,16384,0.008479110896587372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,51200,0.04016000032424927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,12288,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,16384,0.020478222105238173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,10240,0.007304888632562425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,12288,0.017856000198258292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,8192,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,10240,0.016899555921554565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,7168,0.006590222318967183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,8192,0.015458666616015963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,6144,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,7168,0.01513777838812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,5120,0.00590844452381134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,6144,0.014140443669425117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,4096,0.00589155571328269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,5120,0.014112888111008538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,3584,0.0058986664646201665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,4096,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,3072,0.005362666729423735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,3584,0.013421333498424955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,2560,0.0059262220230367445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,3072,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,2048,0.005229333208666907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,2560,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,1536,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,2048,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,1024,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,1536,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,768,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,1024,0.011239111423492432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,512,0.0031724443866146933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,768,0.011728888584507836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,512,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,256,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,256,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,128,0.002551111082235972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,128,0.011367110742463006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,64,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,64,0.011028444601429833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,128,32,0.002491555487116178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,128,32,0.011029333704047732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,65536,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,65536,0.04903555578655667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,51200,0.01035733355416192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,16384,0.007920888562997183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,51200,0.040778666734695435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,12288,0.0070186663005087115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,12288,0.018253333038753934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,16384,0.02023555503951179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,10240,0.006966222491529252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,8192,0.006647999915811751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,8192,0.01588000026014116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,7168,0.006686222222116258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,10240,0.01753866672515869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,7168,0.01520088811715444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,6144,0.005927111125654644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,5120,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,6144,0.014863111906581454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,4096,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,5120,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,3584,0.005623999983072281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,4096,0.013493333425786761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,3072,0.005630222045713001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,3584,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,3072,0.012826667063766055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,2560,0.005653333332803514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,2048,0.005312888986534542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,2560,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,1536,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,2048,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,1536,0.011870221959220039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,1024,0.003988444391224119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,768,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,1024,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,512,0.0032506665835777917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,768,0.011821333732869891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,512,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,256,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,128,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,256,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,64,0.002564444517095884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,128,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,64,32,0.0027520000520679685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,64,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,65536,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,64,32,0.01073688848151101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,51200,0.01036000003417333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,65536,0.04932266804907057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,16384,0.007658667034573025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,16384,0.020612445142534044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,12288,0.007301333049933116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,51200,0.04070933328734504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,12288,0.01792622274822659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,10240,0.007000889215204451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,8192,0.006346666564544042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,10240,0.016921778519948322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,7168,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,8192,0.015554666519165039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,6144,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,7168,0.015319999721315173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,5120,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,6144,0.014502222339312235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,4096,0.005024000174469418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,5120,0.01421511173248291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,3584,0.005983110931184556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,4096,0.013439999686347114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,3584,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,3072,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,2560,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,3072,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,2048,0.005704000178310607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,2560,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,1536,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,2048,0.012061333490742577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,1024,0.003900444342030419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,1536,0.012050666742854647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,768,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,1024,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,512,0.0031519999934567344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,768,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,512,0.011510222322411008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,256,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,128,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,128,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,64,0.0027395555128653846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,64,0.010784889260927836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,32,0.0025173332542181015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,17,32,32,0.010699555277824402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,17,32,256,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,51200,0.49680354860093856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,51200,0.9523039923773872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,16384,0.3392515447404649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,16384,0.16883999771542021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,16384,0.25088622834947377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,12288,0.22644710540771484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,12288,0.1303608947330051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,51200,0.7183128992716471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,10240,0.18961689207288954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,12288,0.19701245095994738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,10240,0.1128044393327501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,8192,0.15268622504340276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,8192,0.09121066994137234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,10240,0.1689128875732422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,7168,0.13442488511403403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,7168,0.10748089022106594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,6144,0.11598222785525852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,8192,0.14092089070214167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,7168,0.1857128938039144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,6144,0.08858044279946221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,5120,0.09818489021725124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,5120,0.07813955677880181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,6144,0.13912622133890787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,4096,0.07819288969039917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,4096,0.0659244457880656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,5120,0.14576799339718288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,3584,0.06957955492867364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,4096,0.10813244183858235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,3584,0.056186669402652316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,3072,0.06033955679999458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,3072,0.051345778836144336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,3584,0.10181955496470134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,2560,0.05103111267089844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,3072,0.08396889103783502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,2560,0.04525599877039591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,2048,0.041640000210867986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,2560,0.07318044371075101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,1536,0.03255555695957608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,2048,0.03740444448259141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,2048,0.07147910859849718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,1536,0.0299288895395067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,1024,0.023289778166347083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,1536,0.054495112763510815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,1024,0.023688889212078516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,1024,0.04522044459978739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,768,0.018837332725524902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,512,0.013237333132161034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,768,0.021036444438828364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,768,0.039988444911109075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,256,0.00903200027015474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,512,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,512,0.03384533193376329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,128,0.00655644428398874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,256,0.015777778294351365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,256,0.03125155634350247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,128,0.01314755529165268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,64,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,65536,32,0.00563466673096021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,65536,128,0.02805866797765096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,64,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,65536,32,0.01622577839427524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,65536,0.5241120126512315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,65536,1.0377911461724174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,51200,0.8022506501939562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,51200,0.40266577402750653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,16384,0.2764640119340685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,16384,0.13833867179022893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,16384,0.21815644370185006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,12288,0.19831466674804688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,12288,0.10712177885903253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,65536,0.7813164393107096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,12288,0.1720942258834839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,10240,0.16760978433820936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,10240,0.0926275584432814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,8192,0.13637510935465494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,8192,0.07561510801315308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,7168,0.11662844816843669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,7168,0.08812889125612046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,10240,0.147889773050944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,8192,0.1231351163652208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,51200,0.6199199888441298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,6144,0.07034755415386625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,5120,0.08642666869693333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,7168,0.1571048895517985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,6144,0.11585688591003418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,5120,0.06408977508544922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,4096,0.07160977522532146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,4096,0.05697866943147448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,5120,0.1387359963523017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,3584,0.0632479985555013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,6144,0.10176622205310398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,3072,0.05262844430075752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,4096,0.07881866561041938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,3584,0.05053066545062595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,2560,0.04469777809249031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,3584,0.08717599842283462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,3072,0.07264710797203912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,2560,0.038966221941841975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,2048,0.038007110357284546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,2560,0.06425422430038452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,2048,0.03226222263442145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,1536,0.029955555995305378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,2048,0.05650933583577474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,1536,0.026241777671708003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,1024,0.020635555187861126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,1536,0.047803554270002574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,1024,0.020413332515292697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,768,0.01587288909488254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,1024,0.038331554995642766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,768,0.034824000464545354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,768,0.01867911054028405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,512,0.01112711098459032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,512,0.01685066686736213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,512,0.03148444493611654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,256,0.007335111498832703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,3072,0.044587555858823985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,128,0.0039013334446483185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,256,0.014645333091417948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,256,0.028892444239722356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,64,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,51200,128,0.023388443721665278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,128,0.013481777575280933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,51200,32,0.003842666745185852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,64,0.014149333039919535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,51200,32,0.01551555593808492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,65536,0.3321422206030952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,65536,0.18255466885036895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,51200,0.2605697843763563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,51200,0.1410213311513265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,16384,0.09805599848429362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,16384,0.05537333422236972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,16384,0.06939822435379028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,12288,0.07050044669045342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,51200,0.1560480064815945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,65536,0.19963110817803276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,12288,0.044673777288860746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,10240,0.06079910861121284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,10240,0.040208889378441706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,8192,0.042577776643964976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,12288,0.057335999276902944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,10240,0.052654223309622876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,8192,0.034381333324644305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,7168,0.03798755672242906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,7168,0.03924355573124356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,8192,0.04706933432155185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,6144,0.03305955396758185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,7168,0.04629955689112345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,5120,0.02852177619934082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,6144,0.03500444359249539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,6144,0.043713778257369995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,4096,0.023306666149033442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,5120,0.03199822372860379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,5120,0.03881511092185974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,3584,0.020656888683636982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,4096,0.027408000495698717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,4096,0.036271999279658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,3584,0.023657777243190344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,3072,0.01834844383928511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,3584,0.03373600045839945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,2560,0.01572177807490031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,3072,0.022456889351209004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,3072,0.03278133273124695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,2048,0.01333866682317522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,2560,0.0210853334930208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,2560,0.030219554901123047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,1536,0.010435555544164447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,2048,0.01903288894229465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,2048,0.028789334826999243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,1024,0.008600888980759515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,1536,0.016211556063758004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,1536,0.02791022260983785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,1024,0.024699555502997503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,768,0.006959999601046245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,1024,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,512,0.004779555731349521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,768,0.01310755560795466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,768,0.022639110684394836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,256,0.0032302221904198327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,512,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,512,0.02029600077205234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,128,0.00313688897424274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,256,0.020396444532606337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,256,0.012143111063374413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,64,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,128,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,16384,128,0.018918222851223417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,16384,32,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,64,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,16384,32,0.012874667015340594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,65536,0.2463137838575575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,65536,0.1592764457066854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,51200,0.194940447807312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,51200,0.1224622196621365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,16384,0.06978488630718656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,16384,0.049025777313444346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,65536,0.16395910580952963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,51200,0.12990932994418675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,12288,0.05102666550212436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,16384,0.06090044313006931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,10240,0.044102221727371216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,12288,0.03988533218701681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,12288,0.05160977774196201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,10240,0.036244445376926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,10240,0.04695999953481886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,8192,0.03749688797526889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,8192,0.03155466583040025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,7168,0.03288266725010342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,8192,0.04256711072391934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,7168,0.03438755538728502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,6144,0.029317332638634577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,7168,0.041871110598246254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,6144,0.03050666716363695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,5120,0.025798221429189045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,6144,0.03870666689342923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,5120,0.02699377801683214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,4096,0.02134399943881565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,5120,0.03717244333691067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,4096,0.0239964442120658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,3584,0.019512888458040025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,4096,0.033929778469933405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,3584,0.022216889593336318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,3584,0.03195022212134467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,3072,0.017229333519935608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,3072,0.030104888810051814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,3072,0.021545777718226116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,2560,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,2560,0.028897778855429754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,2048,0.010571555958853828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,2560,0.01924800044960446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,2048,0.017027555240525138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,2048,0.028975112570656672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,1536,0.009653333160612319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,1536,0.02569688856601715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,1536,0.01609422266483307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,1024,0.008025777836640676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,1024,0.02235466738541921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,1024,0.013162666724787818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,768,0.0064071110553211635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,768,0.020410666863123577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,768,0.012801777985360889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,512,0.004277333203289243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,512,0.01994755533006456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,512,0.01273333364062839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,256,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,256,0.018992000155978732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,128,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,256,0.012048888537618848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,128,0.01069866700304879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,64,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,12288,128,0.019336000084877014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,12288,32,0.003136000078585413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,64,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,65536,0.20881777339511445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,12288,32,0.012515555653307172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,65536,0.1500071154700385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,51200,0.16608088546329075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,51200,0.11762844191657172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,65536,0.1506755616929796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,16384,0.06506666872236463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,16384,0.04732444551255968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,51200,0.12013778421613906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,12288,0.045824001232783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,16384,0.05721600188149346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,12288,0.038776887787712946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,10240,0.03907288776503669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,12288,0.04944444364971585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,10240,0.034620443979899086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,8192,0.0319199992550744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,10240,0.04502844479348925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,8192,0.040622221099005804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,7168,0.029531555043326482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,7168,0.030557334423065186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,7168,0.0396382212638855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,8192,0.03003466791576809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,6144,0.02621600031852722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,6144,0.02607111136118571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,5120,0.02313333253065745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,6144,0.03683111071586609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,5120,0.02366222275627984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,5120,0.035042666726642184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,4096,0.0188400000333786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,4096,0.03194311261177063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,3584,0.015071110592948066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,4096,0.021643555826610986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,3584,0.0199608885579639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,3072,0.013184888495339287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,3584,0.031630221340391375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,3072,0.019048889478047688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,3072,0.03049422303835551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,2560,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,2560,0.028584000137117174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,2048,0.010072000324726105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,2560,0.017920000685585868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,2048,0.016120000018013846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,2048,0.028379556205537584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,1536,0.009054222040706212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,1536,0.02542488939232296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,1024,0.0069822221994400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,1536,0.014955555399258932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,1024,0.012991111311647626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,1024,0.0229768885506524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,768,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,768,0.012753778033786349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,768,0.022085333863894146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,512,0.004642666627963384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,512,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,512,0.02165066699186961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,256,0.003911111089918348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,256,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,256,0.020998222960366145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,128,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,128,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,10240,128,0.019325332509146798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,64,0.0029120000286234748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,64,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,10240,32,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,10240,32,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,65536,0.16316266854604086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,65536,0.10055466492970784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,65536,0.14451911714341906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,51200,0.12950311766730413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,51200,0.08010933134290907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,16384,0.0516844458050198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,16384,0.0354364448123508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,51200,0.11738310919867621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,12288,0.032595554987589516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,16384,0.054810666375690036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,12288,0.029823111163245306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,10240,0.027871999475691054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,12288,0.048162665632036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,10240,0.026496888862715826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,8192,0.02334755493534936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,10240,0.04407199886110094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,8192,0.022442667020691767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,7168,0.020945777495702107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,8192,0.03945511248376634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,7168,0.03715022073851691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,6144,0.018396443790859647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,7168,0.023630221684773762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,6144,0.02203377750184801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,6144,0.03578133384386698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,5120,0.01587288909488254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,4096,0.013192888763215808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,5120,0.020253333780500624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,5120,0.03436177637841966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,3584,0.012042666474978128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,4096,0.01887022289964888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,4096,0.03252799974547492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,3072,0.010461333725187512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,3584,0.018547554810841877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,3584,0.030510223574108545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,3072,0.01697688963678148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,2560,0.009305777649084726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,3072,0.030184000730514526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,2048,0.008038221961922115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,2560,0.015857777661747403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,2560,0.0281644430425432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,1536,0.00692622239391009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,2048,0.027120888233184814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,2048,0.014839111102951897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,1024,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,1536,0.025091555383470323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,1536,0.013330666555298699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,768,0.004280888785918553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,1024,0.012500444220172035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,1024,0.021683555510309007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,512,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,768,0.011867555479208628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,768,0.021708443760871887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,256,0.0032604444358083936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,512,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,512,0.020442666278945077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,256,0.02014755540423923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,256,0.01180355581972334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,128,0.011210666762457954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,64,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,8192,128,0.019653333557976615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,32,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,64,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,8192,32,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,65536,0.15257510874006483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,65536,0.09191022316614787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,51200,0.11469688680436875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,8192,128,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,51200,0.07335199912389119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,16384,0.04282844397756788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,65536,0.14339999357859293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,51200,0.11634488900502522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,16384,0.03327200147840712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,16384,0.05454666746987236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,12288,0.031826668315463595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,10240,0.027973333994547527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,12288,0.02805244591501024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,12288,0.04691911074850294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,8192,0.02335022224320306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,10240,0.024456888437271118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,10240,0.0435751113626692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,7168,0.021293333835071985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,8192,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,8192,0.03829511006673177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,6144,0.016718222035302054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,7168,0.03743733300103082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,7168,0.02348088887002733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,6144,0.02120622164673275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,6144,0.0352862212393019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,5120,0.01493866741657257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,5120,0.020278222031063504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,5120,0.033189333147472806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,4096,0.012458666331238218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,4096,0.018206222189797294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,3584,0.011506666739781698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,3584,0.017536888519922893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,3584,0.03050400151146783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,3072,0.009704888694816166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,3072,0.016553777787420485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,4096,0.0322106679280599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,2560,0.009278222090668148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,3072,0.029500444730122883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,2560,0.027783112393485174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,2560,0.015807999504937064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,2048,0.0076657773719893555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,2048,0.01440711153878106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,2048,0.026781333817376032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,1536,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,1536,0.012825777961148156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,1536,0.025080889463424683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,1024,0.0052977779673205475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,1024,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,1024,0.02165511084927453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,768,0.0039057777159743835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,768,0.02200977835390303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,768,0.011786667009194693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,512,0.003488000068399641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,512,0.021338666478792827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,512,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,256,0.0032026666320032547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,256,0.010778666370444827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,256,0.020967110991477966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,128,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,128,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,7168,128,0.018276444739765592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,64,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,7168,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,64,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,7168,32,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,65536,0.12899911403656006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,51200,0.09649155537287395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,65536,0.08405688736173843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,51200,0.06689511405097114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,16384,0.03674222363366021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,65536,0.1425724426905314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,16384,0.03137955400678847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,51200,0.11695999569363064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,12288,0.028107557031843398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,16384,0.053935110569000244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,10240,0.02494488822089301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,12288,0.025574222207069397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,12288,0.04704888992839389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,10240,0.023178666830062866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,8192,0.020824889341990154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,10240,0.04311733444531759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,8192,0.02033422225051456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,7168,0.018989332848125033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,8192,0.03814311159981622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,6144,0.017122666041056316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,7168,0.022720888257026672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,7168,0.037774221764670476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,5120,0.015257777439223396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,6144,0.03466488917668661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,5120,0.03263288736343384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,5120,0.019712888532214694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,4096,0.011142222417725457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,4096,0.030942221482594807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,4096,0.01794311073091295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,3584,0.010375111467308467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,3584,0.02940622303220961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,3072,0.009272000028027428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,3584,0.01733688844574822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,3072,0.016232000456915963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,3072,0.0295688874191708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,2560,0.008054222497675154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,6144,0.020611556039916146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,2048,0.007627555893527136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,2560,0.02769333289729224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,2560,0.016166221764352586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,1536,0.006245333287451003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,2048,0.014024888475735983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,2048,0.025841777523358662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,1024,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,1536,0.01315822203954061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,1536,0.02334844403796726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,1024,0.01204533295498954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,768,0.0036106668412685394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,1024,0.022257778379652236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,512,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,768,0.011915555430783166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,768,0.020351111888885498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,256,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,512,0.020310221446885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,512,0.012123555772834353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,128,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,256,0.010999999940395355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,256,0.01891377733813392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,128,0.011596444580290051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,6144,128,0.01720711092154185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,64,0.002700444517864121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,6144,32,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,64,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,6144,32,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,65536,0.07678755786683825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,65536,0.14278844992319742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,65536,0.10706755850050186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,51200,0.08719733026292588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,51200,0.06035288837220934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,16384,0.03477511141035292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,16384,0.028613332245084975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,51200,0.11657244629330105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,12288,0.02602488961484697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,16384,0.05367822117275662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,12288,0.023381332556406658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,10240,0.02199111051029629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,12288,0.046201778782738574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,10240,0.042188445727030434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,8192,0.019394665956497192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,10240,0.022050667140218947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,8192,0.019319999549123976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,7168,0.017112889223628573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,8192,0.038365334272384644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,7168,0.02167911165290409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,7168,0.0374951097700331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,6144,0.015627556376987033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,6144,0.019931554794311523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,6144,0.03456799851523505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,5120,0.014122666584120857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,5120,0.01852888862291972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,4096,0.010670222342014313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,5120,0.0336844457520379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,4096,0.017279111676745944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,3584,0.009845332966910468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,4096,0.03179555469089084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,3584,0.016207110550668504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,3584,0.030525333351559106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,3072,0.009134222235944536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,3072,0.0284862220287323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,3072,0.015118221441904703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,2560,0.007619555625650618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,2560,0.02548800077703264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,2560,0.014523555835088095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,2048,0.00674933319290479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,2048,0.013524444566832649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,2048,0.024000888069470722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,1536,0.00628355559375551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,1536,0.01238222254647149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,1536,0.023982221881548565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,1024,0.004218666503826777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,1024,0.02067733307679494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,1024,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,768,0.01204533295498954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,768,0.021672889590263367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,512,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,512,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,512,0.02130844526820713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,256,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,768,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,256,0.011693333586057028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,256,0.018962666392326355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,128,0.011364444262451597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,5120,128,0.01722133325205909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,32,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,64,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,5120,32,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,65536,0.08703377511766221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,65536,0.06975911060969035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,51200,0.06705866919623481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,51200,0.05673866801791721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,65536,0.14261689451005724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,16384,0.028189334604475234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,5120,64,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,16384,0.025641777449183997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,12288,0.02093688812520769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,51200,0.11572444438934326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,12288,0.022237333986494277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,12288,0.04513066675927904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,10240,0.018195556269751657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,10240,0.02021511064635383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,8192,0.01536355581548479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,10240,0.04202488727039761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,8192,0.038154668278164335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,8192,0.01830400029818217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,7168,0.014256000518798828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,7168,0.02029244436158074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,7168,0.037485334608289934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,16384,0.053415109713872276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,6144,0.01871200071440803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,5120,0.012014221813943652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,6144,0.03554666704601712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,5120,0.017674666312005784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,5120,0.03276888860596551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,4096,0.011911110745535957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,4096,0.016143111719025504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,4096,0.03179733289612664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,3584,0.010817777779367236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,3584,0.014730667074521383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,6144,0.013169777890046438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,3072,0.009704888694816166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,3584,0.029931555191675823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,3072,0.01408266690042284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,2560,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,3072,0.02719733284579383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,2560,0.02574577762020959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,2048,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,2048,0.01311644415060679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,2048,0.02421866688463423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,1536,0.004540444248252445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,1536,0.012095111111799875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,1536,0.022667555345429316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,1024,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,2560,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,1024,0.011993777420785693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,768,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,1024,0.020966221888860066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,768,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,512,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,768,0.01961599952644772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,512,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,512,0.01165688865714603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,256,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,256,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,256,0.018584888842370775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,128,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,128,0.010388444695207808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,4096,128,0.01889600025282966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,64,0.0027031112048361036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,4096,32,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,64,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,4096,32,0.010679110884666443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,65536,0.07602844635645549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,65536,0.0671795540385776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,51200,0.05869688590367635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,65536,0.14185955789354113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,51200,0.05463822351561653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,16384,0.024683554967244465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,16384,0.024339555038346186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,51200,0.11548356215159099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,12288,0.018555555078718398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,16384,0.05313422282536825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,12288,0.021153777837753296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,12288,0.04491200049718221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,10240,0.01627200014061398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,10240,0.019991111424234178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,10240,0.04237777656979031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,8192,0.014202665951516895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,8192,0.03829066620932685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,8192,0.018631110588709515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,7168,0.014446222119861178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,6144,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,7168,0.019711999429596793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,7168,0.03665688965055678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,5120,0.011054221954610614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,6144,0.03424533208211263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,6144,0.018552889426549275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,4096,0.010913777682516309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,5120,0.01696888936890496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,5120,0.03196444445186191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,3584,0.010260444548394945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,4096,0.031585776143603854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,4096,0.015494222442309061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,3072,0.009290667043791877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,3584,0.014508444401952954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,3584,0.029523554775449965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,2560,0.00676711110605134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,3072,0.026118222210142348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,3072,0.013934221532609729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,2048,0.005383111122581694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,2560,0.013434666726324292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,2560,0.02511022157139248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,1536,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,2048,0.01275199982855055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,2048,0.02405511173937056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,1024,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,1536,0.011824000212881299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,1536,0.022652443912294176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,1024,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,768,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,1024,0.020652444826232064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,512,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,768,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,768,0.02058577703105079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,256,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,512,0.01108000013563368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,512,0.020630222227838304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,128,0.0028435554769304064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,256,0.018872888551818002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,256,0.010586666564146677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,64,0.002546666603949335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,128,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3584,128,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3584,32,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,64,0.01035911093155543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,65536,0.06269066863589816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3584,32,0.010682666467295753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,65536,0.06446311208936903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,51200,0.05073600014050802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,65536,0.1417644421259562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,51200,0.052557332648171314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,16384,0.021280889709790547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,51200,0.11555466387006973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,16384,0.024007999234729346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,16384,0.05344177616967095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,12288,0.016551110479566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,12288,0.020844444632530212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,12288,0.04594399862819248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,10240,0.01501511037349701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,10240,0.042261332273483276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,10240,0.019519999623298645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,8192,0.013047999805874295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,8192,0.03830400109291077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,7168,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,7168,0.018266666266653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,7168,0.03664977682961358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,6144,0.010944888823562197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,6144,0.017325333423084684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,6144,0.03538577755292257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,5120,0.010233778092596266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,8192,0.017493334081437852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,5120,0.016420443852742512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,4096,0.009908444351620143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,5120,0.03289688958062066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,4096,0.014575110541449653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,3584,0.009172444542249044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,4096,0.028888887829250757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,3584,0.014134221606784396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,3072,0.008240888516108194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,3584,0.02780888809098138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,3072,0.013216888739003075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,2560,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,3072,0.026160000099076167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,2560,0.012826667063766055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,2048,0.00489066665371259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,2560,0.024737776981459722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,2048,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,1536,0.004287999951177173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,2048,0.02437155610985226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,1536,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,1536,0.022633777724372015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,1024,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,1024,0.02132533325089349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,768,0.0036755556033717263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,768,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,768,0.020652444826232064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,1024,0.0038728887836138406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,512,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,512,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,512,0.020271999968422782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,256,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,256,0.020266667008399963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,256,0.010560000108347999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,128,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,64,0.002524444419476721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,3072,128,0.01828533411026001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,3072,32,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,64,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,65536,0.056223114331563316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,3072,32,0.01112711098459032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,51200,0.04607289036115011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,65536,0.062078224288092725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,51200,0.05083022183842129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,65536,0.14222578207651773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,16384,0.019835554891162448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,51200,0.11521422863006592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,12288,0.01513777838812934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,16384,0.05232977867126465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,12288,0.02030933399995168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,10240,0.014084445105658637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,12288,0.04582666688495212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,10240,0.018943111101786297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,8192,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,10240,0.041598220666249595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,8192,0.01756266587310367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,8192,0.03746044304635789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,7168,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,7168,0.036658667855792575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,7168,0.01795733306143019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,6144,0.010208889014191097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,6144,0.03512444429927402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,16384,0.023726221587922838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,6144,0.01704088846842448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,5120,0.00942666663063897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,4096,0.009200889203283522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,5120,0.03157866663402981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,4096,0.014139556222491793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,3584,0.008681777450773451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,4096,0.028627554575602215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,3584,0.013460444079505073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,3072,0.007794666621420119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,3584,0.027463111612531874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,3072,0.013181333740552267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,2560,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,3072,0.026125333375400964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,2560,0.012806221842765808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,2560,0.02513155506716834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,2048,0.004956444518433677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,5120,0.014872888724009195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,2048,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,2048,0.0251528885629442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,1536,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,1536,0.022699554761250813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,1536,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,1024,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,1024,0.022337777747048274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,768,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,768,0.01202311118443807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,1024,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,512,0.003301333429084884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,768,0.021783111823929682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,512,0.021391999390390184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,512,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,256,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,256,0.020682666036817763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,256,0.01075111081202825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,128,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2560,128,0.019334221879641216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,64,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,128,0.010433778166770935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2560,32,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,64,0.010773333410422007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2560,32,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,65536,0.04563911093605889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,65536,0.05968533621893989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,51200,0.0380951099925571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,65536,0.14208622773488364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,51200,0.049160887797673546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,16384,0.015182221929232279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,51200,0.11571644412146674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,16384,0.022830223043759663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,16384,0.053060445520612926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,12288,0.013033777475357056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,12288,0.02035555574629042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,12288,0.045560889773898654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,10240,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,10240,0.04192355606291029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,10240,0.01863911085658603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,8192,0.010402667025725046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,8192,0.017055999901559617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,7168,0.010737777584128909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,7168,0.016200888488027785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,7168,0.03617422117127313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,6144,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,6144,0.015783111254374187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,8192,0.037689778539869524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,5120,0.011234666738245221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,6144,0.03225422236654494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,4096,0.009863110880057016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,5120,0.030321776866912842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,4096,0.01351999988158544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,3584,0.00720355576939053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,3584,0.013146667016877068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,3584,0.02680355476008521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,3072,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,3072,0.013056000073750814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,5120,0.014461333552996317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,3072,0.025775111383861963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,2560,0.005678222411208683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,4096,0.028505778974956934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,2560,0.012617777619096967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,2048,0.004912888838185204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,2560,0.025623111261261836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,1536,0.004600000050332812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,2048,0.012175111307038201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,2048,0.024146666129430134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,1024,0.003916444463862313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,1536,0.021713778376579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,1536,0.012514666550689273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,768,0.003612444632583194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,1024,0.020769778225156996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,768,0.019634667370054457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,512,0.0032746667663256326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,768,0.011972444752852121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,512,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,256,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,512,0.021336888273557026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,256,0.010880000061459012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,256,0.021014221840434607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,128,0.010455110834704505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,2048,128,0.019339554839664035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,64,0.0029848888516426086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,1024,0.01148977792925305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,2048,32,0.0025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,65536,0.03610310951868693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,64,0.010966222319338055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,2048,32,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,65536,0.05778400103251139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,51200,0.029139555162853662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,65536,0.1421075529522366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,51200,0.047242667939927846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,16384,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,51200,0.11537333329518636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,16384,0.022455111145973206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,12288,0.015875554747051664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,16384,0.05311288767390781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,12288,0.019885333047972787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,10240,0.014212444424629211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,12288,0.045984887414508395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,10240,0.017627555463049147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,8192,0.012500444220172035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,8192,0.015848888291252982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,8192,0.0369742214679718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,7168,0.011494222614500256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,7168,0.015804444750150044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,7168,0.03394577900568644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,6144,0.01069155583779017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,10240,0.04197777642144097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,6144,0.014843554960356818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,5120,0.009312888814343346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,6144,0.03232266505559286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,5120,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,4096,0.008622222476535374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,5120,0.03122666809293959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,4096,0.01351999988158544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,3584,0.007296889192528195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,4096,0.02882755464977688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,3584,0.013168888787428537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,3584,0.028223110569847956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,3072,0.006423111177153057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,3072,0.025721778472264607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,3072,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,2560,0.005600888695981767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,2560,0.024706666668256123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,2560,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,2048,0.005672889037264719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,2048,0.01240711079703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,1536,0.004541333350870344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,1536,0.012320888539155325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,1536,0.023730667101012334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,1024,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,1024,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,1024,0.02237422267595927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,768,0.0036115555299652945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,2048,0.02344266739156511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,768,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,768,0.01926666663752662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,512,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,512,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,512,0.021340444684028625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,256,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,256,0.0114702218108707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,256,0.018636445204416912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,128,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,128,0.010335110955768162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1536,128,0.019319110446506076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,64,0.002557333351837264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1536,32,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,64,0.011023999916182624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1536,32,0.010398222340477837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,65536,0.02567377852069007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,65536,0.05602044529385037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,51200,0.021838222940762837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,65536,0.14195111062791613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,16384,0.014104889498816596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,51200,0.0448293321662479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,51200,0.11444800429873997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,12288,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,16384,0.02240444388654497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,16384,0.05265600151485867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,10240,0.010695111420419482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,12288,0.01862844493654039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,12288,0.045540445380740695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,8192,0.01016977760526869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,10240,0.017177777157889474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,10240,0.04045244389110141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,7168,0.008993777963850234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,8192,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,8192,0.03615555498335097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,6144,0.008360000120268928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,7168,0.015174221661355762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,7168,0.034683555364608765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,5120,0.008004444340864817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,6144,0.014771555860837301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,6144,0.03183111217286851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,4096,0.007253333098358578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,5120,0.014481777946154276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,5120,0.030293332205878362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,3584,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,4096,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,4096,0.02903111113442315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,3072,0.006197333335876465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,3584,0.02643999954064687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,3072,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,3072,0.025455110602908667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,2560,0.005569777968857024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,2560,0.012464000119103326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,2560,0.025721778472264607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,2048,0.006573333508438534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,2048,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,2048,0.023044443792766992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,1536,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,1536,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,1024,0.004602666530344221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,1536,0.0236835562520557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,1024,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,3584,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,768,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,1024,0.020643555455737643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,768,0.011381333072980246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,512,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,768,0.019613333874278598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,512,0.011341333389282227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,512,0.01925244430700938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,256,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,256,0.010763555765151978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,256,0.020301333732075162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,128,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,128,0.011137777732478248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,1024,128,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,64,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,1024,32,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,64,0.01035288886891471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,65536,0.021809778279728357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,1024,32,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,65536,0.054690665668911405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,51200,0.02713866697417365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,65536,0.14115822315216064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,51200,0.04416888952255249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,16384,0.012166221936543783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,16384,0.020657777786254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,16384,0.05112444361050924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,12288,0.010446222292052375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,12288,0.018211555149820115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,12288,0.043460445271597974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,10240,0.009353777600659264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,51200,0.11454044448004828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,10240,0.01681422193845113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,8192,0.009383999639087254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,10240,0.04008177916208903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,8192,0.015805333852767944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,7168,0.008631111019187504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,8192,0.03594311078389486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,7168,0.01516355574131012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,7168,0.03394577900568644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,6144,0.00795733349190818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,6144,0.014771555860837301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,6144,0.03155022197299533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,5120,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,5120,0.014116444521480136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,5120,0.030223998758527968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,4096,0.00702044450574451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,4096,0.013420444395807056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,4096,0.028157333532969158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,3584,0.006663110935025745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,3584,0.013048888908492194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,3584,0.02765688962406582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,3072,0.0063075555695427795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,3072,0.025729777084456548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,3072,0.01273333364062839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,2560,0.006233777850866318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,2560,0.012502222425407834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,2560,0.024658666716681585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,2048,0.004920888692140579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,2048,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,2048,0.024041778511471216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,1536,0.004934222333961063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,1536,0.021707556313938562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,1536,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,1024,0.0041991112132867174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,1024,0.011705777711338468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,1024,0.020298666424221463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,768,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,768,0.010737777584128909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,768,0.020583111378881667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,512,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,512,0.01926577753490872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,512,0.011295999917719098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,256,0.011028444601429833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,256,0.01829955478509267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,128,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,128,0.010049777726332346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,768,128,0.0189057770702574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,64,0.0028311111446883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,768,32,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,64,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,65536,0.021963554951879714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,65536,0.053655998574362866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,65536,0.14062044355604383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,51200,0.019137778215938143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,51200,0.04415555463896858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,51200,0.11453155676523845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,16384,0.010386666489972008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,768,32,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,12288,0.008658666577604082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,16384,0.051390220721562706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,12288,0.017838221457269456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,10240,0.007624000310897827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,12288,0.042824000120162964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,16384,0.0206675562593672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,10240,0.016795555750528973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,8192,0.007927999728255803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,10240,0.040072888135910034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,7168,0.0076044441925154785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,8192,0.01550222271018558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,8192,0.03499911228815714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,7168,0.01514488955338796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,6144,0.007334222396214803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,7168,0.033922665648990206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,6144,0.014491554763582019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,5120,0.006883555402358373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,5120,0.014175999495718213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,5120,0.02951999836497837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,4096,0.007066667079925537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,4096,0.013408888545301227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,4096,0.027842667367723253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,3584,0.006732444382376141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,6144,0.03164088726043701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,3584,0.013117333253224691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,3072,0.006495111104514863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,3584,0.027090667022599116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,3072,0.012763555679056378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,2560,0.006301333506902059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,3072,0.02611911131276025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,2560,0.012375999655988483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,2048,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,2560,0.0247653325398763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,2048,0.011844444606039258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,2048,0.023748444186316595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,1536,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,1536,0.021631111701329548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,1024,0.003932444585694207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,1024,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,1024,0.020601777566803824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,768,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,768,0.010682666467295753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,768,0.019278221660190158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,512,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,1536,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,512,0.01072622256146537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,256,0.003126222226354811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,512,0.020012444920010038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,256,0.010454222559928894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,256,0.0192239996459749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,128,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,512,128,0.01723022262255351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,128,0.010021333065297868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,64,0.010023111270533668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,32,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,512,32,0.01073333372672399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,65536,0.015503111812803479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,65536,0.049977779388427734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,512,64,0.0026497778793176017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,65536,0.1404942274093628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,51200,0.013080000049538083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,51200,0.04231733414861891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,51200,0.11297510729895698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,16384,0.00867911097076204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,16384,0.01994044416480594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,16384,0.049313777022891574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,12288,0.008031999899281396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,12288,0.04215911030769348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,10240,0.0075253330998950535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,10240,0.016555555992656283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,10240,0.039108445247014366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,8192,0.007241778075695038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,8192,0.015449777245521545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,7168,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,7168,0.014930667148696052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,8192,0.03569599986076355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,6144,0.006606222026877933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,12288,0.017880888448821176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,7168,0.03292355603641934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,5120,0.00620888877246115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,6144,0.03190222382545471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,5120,0.013902222116788229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,6144,0.014894222219785055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,4096,0.006311111152172089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,5120,0.029885331789652508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,3584,0.006190222170617845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,4096,0.027856889698240492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,4096,0.013546666337384118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,3072,0.005941333456171884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,3584,0.012803555362754397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,3584,0.026308443811204698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,2560,0.006275555739800136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,3072,0.013100444442696042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,3072,0.026134222745895386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,2048,0.0052897776994440295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,2560,0.012396444049146442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,2560,0.025761778155962627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,1536,0.004563555535342959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,2048,0.02299555473857456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,2048,0.012073777616024017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,1024,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,1536,0.021959111094474792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,1536,0.01219288922018475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,1024,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,768,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,1024,0.0203155560625924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,512,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,768,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,768,0.022056889202859666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,512,0.010716444088353051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,256,0.003169777699642711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,512,0.019906666543748643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,256,0.01959111127588484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,128,0.0028213332924577924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,128,0.010055999788973067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,256,128,0.017941332525677152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,64,0.0025155554629034465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,64,0.010102222363154093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,256,32,0.002528000002106031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,256,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,65536,0.013893332746293811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,256,32,0.010779555473062726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,65536,0.04934755629963345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,51200,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,65536,0.13939021693335638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,51200,0.04090933336151971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,51200,0.11198400126563178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,16384,0.020415999823146395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,16384,0.049366222487555615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,12288,0.007345777418878343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,12288,0.01793155570824941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,16384,0.008002666963471307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,10240,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,12288,0.04369866516855028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,10240,0.017314665847354464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,8192,0.0070062221752272705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,10240,0.03881511092185974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,8192,0.01586666703224182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,7168,0.006666666517655055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,8192,0.035066667530271746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,7168,0.015421332584487068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,6144,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,7168,0.03402488761478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,6144,0.014847111370828418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,5120,0.006302222195598815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,6144,0.03193599979082743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,5120,0.014176888598336114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,4096,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,5120,0.030015110969543457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,4096,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,3584,0.005964444329341252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,4096,0.028194665908813477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,3584,0.013493333425786761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,3072,0.005956444475385878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,3584,0.02751022246148851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,3072,0.013220444321632385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,3072,0.026164443956481084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,2560,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,2560,0.025754666990704004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,2560,0.01240711079703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,2048,0.006018666757477655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,2048,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,2048,0.02514222264289856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,1536,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,1536,0.012162666353914471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,1536,0.022704000274340313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,1024,0.004071111066473855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,768,0.003526222374704149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,1024,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,1024,0.02130844526820713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,768,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,512,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,768,0.02197866638501485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,256,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,512,0.011121778024567498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,512,0.020284444093704224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,128,0.0025804444319672054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,256,0.02014488809638553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,256,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,16,128,128,0.018579555882347953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,128,32,0.002530666689078013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,64,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,65536,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,65536,0.048991110589769155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,51200,0.00981600003110038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,51200,0.04071377714474996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,16384,0.006993778049945831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,128,0.010018666585286459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,16384,0.020607999629444547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,128,32,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,10240,0.006109333286682765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,12288,0.01832177738348643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,8192,0.006679111056857639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,10240,0.016936888297398884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,7168,0.00629688882165485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,8192,0.015909334023793537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,6144,0.006311111152172089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,7168,0.0151582227812873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,5120,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,6144,0.014840889308187695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,5120,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,4096,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,3584,0.006039999839332368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,12288,0.006021333237489064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,4096,0.013551111022631327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,3584,0.013452444639470843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,3072,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,2560,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,2048,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,2560,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,2048,0.012368000215954252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,1536,0.004239999999602635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,1536,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,1024,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,3072,0.005615111026499007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,768,0.0032702222880389956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,1024,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,512,0.0033111110743549135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,512,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,768,0.011175110936164856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,256,0.002942222274012036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,128,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,128,0.011123555401961008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,256,0.010430222584141625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,64,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,64,32,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,32,0.011069333387745751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,64,64,0.010075555907355415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,65536,0.00942755573325687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,51200,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,51200,0.04073511229621039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,65536,0.04874755607710945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,16384,0.020388444264729817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,12288,0.006360000206364526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,12288,0.018079110317760043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,10240,0.006317333214812809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,16384,0.006587555425034628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,8192,0.005596444424655702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,10240,0.01667022208372752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,8192,0.01588622232278188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,7168,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,6144,0.005744888964626525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,7168,0.015170666906568738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,5120,0.006241777704821692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,5120,0.01425155500570933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,6144,0.014578666951921252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,4096,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,3584,0.00572533326016532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,3584,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,3072,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,4096,0.01331377774477005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,2560,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,3072,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,2048,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,2560,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,2048,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,1536,0.004266666869322459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,1024,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,1536,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,768,0.003528888854715559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,1024,0.011184888581434885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,768,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,512,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,256,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,256,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,128,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,128,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,64,0.0027093332674768236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,32,0.0026568888376156488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,64,0.011041777829329172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,16,32,512,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,16,32,32,0.010422222316265106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,16384,0.3363164530860053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,51200,0.4965946409437392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,16384,0.16821511586507162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,51200,0.9525563981797961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,12288,0.22645778126186797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,10240,0.18951466348436144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,12288,0.13076089488135442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,8192,0.11370578077104355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,10240,0.11248177952236599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,8192,0.15256888336605495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,7168,0.13430044386121961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,6144,0.11597777737511529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,7168,0.0961466696527269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,5120,0.09818222125371297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,6144,0.08868622117572361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,5120,0.07774488793479072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,4096,0.07794488800896539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,3584,0.06935022274653117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,3072,0.06020355224609375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,3584,0.056815114286210805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,2560,0.05074400040838453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,3072,0.0506915549437205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,2048,0.04183200001716614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,2048,0.04040711124738058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,1536,0.03276888860596551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,1536,0.03283022178543939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,1024,0.02330044408639272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,4096,0.06207378043068779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,1024,0.024967110819286768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,768,0.018519111805491976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,512,0.013334222137928009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,768,0.02202933364444309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,256,0.008736000292830998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,2560,0.046691556771596275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,128,0.005933333188295364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,256,0.01590755581855774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,64,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,128,0.01443911095460256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,64,0.015470221638679504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,65536,32,0.005611555443869696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,32,0.015861334072219003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,65536,512,0.01865244408448537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,65536,0.5223964585198296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,65536,1.0337147182888455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,51200,0.4018293221791585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,51200,0.7986204359266492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,16384,0.27489866150750053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,16384,0.13760622342427573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,12288,0.19803199503156874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,10240,0.16596710681915283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,12288,0.10703821976979573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,8192,0.13610577583312988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,10240,0.09256800015767415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,8192,0.09189244773652817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,7168,0.1179315514034695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,7168,0.07935644520653619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,6144,0.10454755359225804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,5120,0.08541955550511678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,4096,0.06973777876959907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,5120,0.06671733326382108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,3584,0.06234044498867459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,4096,0.053767999013264976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,3072,0.05297688974274529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,6144,0.07432177993986341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,3584,0.04785155587726169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,2560,0.04478755593299866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,3072,0.043191111750072904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,2048,0.0370399985048506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,2560,0.04073777794837952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,1536,0.02906844351026747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,1024,0.020621332857343886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,2048,0.03527822097142538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,768,0.0158915552828047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,1536,0.02843822373284234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,512,0.011716444459226398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,768,0.020184000333150227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,256,0.007425777614116669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,512,0.017065778374671936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,128,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,1024,0.023099554909600153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,256,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,128,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,32,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,64,0.014460444450378418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,51200,32,0.014480888843536377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,65536,0.33130399386088055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,65536,0.1845982207192315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,51200,0.2606080108218723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,51200,0.14081066184573704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,16384,0.0968080030547248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,12288,0.07066755824618869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,16384,0.054786665572060474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,51200,64,0.003755555384688907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,10240,0.06083289119932386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,8192,0.042944888273874916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,12288,0.04477333360248142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,7168,0.03795377744568719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,8192,0.0414613319767846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,6144,0.03276355399025811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,7168,0.037852442926830716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,5120,0.028605331977208454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,6144,0.034619556532965765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,5120,0.03212799959712558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,4096,0.023390221926901076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,3584,0.020949333906173706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,4096,0.026230222649044458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,3072,0.018202667435010273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,3584,0.023722666833135817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,2560,0.015895111693276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,3072,0.022413333257039387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,2048,0.013058666553762225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,2560,0.021216000119845074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,1536,0.010676444404655032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,2048,0.019911110401153564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,1024,0.008100444244013892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,1536,0.01722044414944119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,10240,0.040629333919949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,768,0.007082666787836287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,512,0.004867555366622077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,1024,0.015132443772421943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,256,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,512,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,128,0.003168888803985384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,256,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,64,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,768,0.014303111367755465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,128,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,16384,32,0.0030533334033356774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,64,0.012143111063374413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,16384,32,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,65536,0.24585066901312935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,51200,0.19460799958970812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,65536,0.15731467141045466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,16384,0.07035911083221436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,12288,0.05113600028885735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,16384,0.04887110988299052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,10240,0.0440248880121443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,12288,0.039903998374938965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,51200,0.12207378281487359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,8192,0.037422223223580256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,10240,0.03596266773011949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,7168,0.03295911020702786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,7168,0.03297955460018582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,6144,0.029167999823888142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,8192,0.042232887612448804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,5120,0.024524443679385718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,6144,0.029283556673261855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,4096,0.019932443896929424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,4096,0.024106666445732117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,3584,0.017639999588330586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,3584,0.021948445174429152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,3072,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,3072,0.021497777766651575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,2560,0.013588444226317935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,2560,0.020098666350046795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,2048,0.011617778076065911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,2048,0.017952000101407368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,5120,0.027499554885758296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,1536,0.009923555784755284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,1024,0.007802666889296637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,1536,0.016541333662139047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,768,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,768,0.013606222139464485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,512,0.004368888835112254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,512,0.012832000023788877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,256,0.0037431112594074676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,1024,0.013815999858909182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,256,0.012726222475369772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,128,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,64,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,128,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,64,0.011615999870830111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,12288,32,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,12288,32,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,65536,0.20661777920193145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,51200,0.1615697807735867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,65536,0.14942222171359593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,16384,0.06364799870385064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,51200,0.11786754926045735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,16384,0.04694044589996338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,12288,0.04641688863436381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,10240,0.03891111082500882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,12288,0.038550221257739596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,8192,0.031722668144438006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,10240,0.03475555446412828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,7168,0.029358221424950495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,8192,0.03182933396763272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,6144,0.02604533400800493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,7168,0.029725333054860432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,6144,0.025792888469166223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,5120,0.022799111074871484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,4096,0.017979555659823947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,5120,0.024162666665183172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,3584,0.016436444388495553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,4096,0.022443556123309668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,3072,0.014498665928840637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,3584,0.019887111253208585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,2560,0.012840000291665396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,3072,0.019528888993793063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,2560,0.018045332696702745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,2048,0.011182222101423474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,1536,0.009398221969604492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,2048,0.016731555263201397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,1024,0.007592889169851939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,1536,0.015604444675975375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,1024,0.013081777426931592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,768,0.013177778157922955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,512,0.004882666799757215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,256,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,512,0.012433778080675336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,128,0.0032311110860771606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,256,0.011536888778209686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,64,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,128,0.012227555943859948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,64,0.011418666276666852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,32,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,65536,0.16315910551283094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,10240,768,0.0057297779454125305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,10240,32,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,51200,0.12974133756425646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,16384,0.05159999926884969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,51200,0.07962133487065633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,65536,0.10123022397359212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,12288,0.03265600072013007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,10240,0.027943111128277246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,16384,0.03516088922818502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,8192,0.023200889428456623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,12288,0.029599110285441082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,7168,0.020815999971495736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,10240,0.026864889595243666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,8192,0.0258240004380544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,6144,0.018383999665578205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,5120,0.015860444969601102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,7168,0.023990222149425085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,4096,0.013176889055305056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,6144,0.02202844454182519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,3584,0.012157333393891653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,5120,0.020985777179400127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,4096,0.019519999623298645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,3072,0.01070755554570092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,3584,0.01761777698993683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,2560,0.009768889182143742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,3072,0.017498667041460674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,2048,0.008406221866607666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,2560,0.01645599967903561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,1536,0.0073724447025193115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,2048,0.015571556157535978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,1024,0.0053342220683892565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,1536,0.013839110732078552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,768,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,1024,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,512,0.003569777641031477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,768,0.012157333393891653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,256,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,512,0.012227555943859948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,128,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,256,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,64,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,128,0.011855111353927188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,8192,32,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,64,0.011504000259770287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,65536,0.1535466644499037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,65536,0.0918088886472914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,51200,0.11442132790883382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,51200,0.07333599858813815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,16384,0.04229244589805603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,8192,32,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,12288,0.0332640012105306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,16384,0.03301422132386102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,10240,0.028844445943832397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,12288,0.028001778655582007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,8192,0.0240346673462126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,10240,0.02423111100991567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,7168,0.022083555658658344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,8192,0.02771200074089898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,6144,0.016920889417330425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,7168,0.022703111171722412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,5120,0.014691554837756686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,6144,0.02221777704026964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,4096,0.012149333126015134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,5120,0.02036977807680766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,3584,0.011227555572986603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,4096,0.01942755613062117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,3072,0.01019377758105596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,3584,0.01754577789041731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,2560,0.009200889203283522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,3072,0.01715555621518029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,2048,0.007992000215583378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,2560,0.015994666351212394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,1536,0.0069679998689227635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,2048,0.015233778291278414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,1024,0.004671999977694618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,1536,0.013647110925780402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,768,0.0036524443162812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,1024,0.012850667039553324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,512,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,768,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,256,0.0032479998966058097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,512,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,128,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,256,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,64,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,128,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,7168,32,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,64,0.011179555621412067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,7168,32,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,65536,0.12948977947235107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,51200,0.09674221939510769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,65536,0.08422933022181193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,16384,0.03705599904060364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,51200,0.06686488787333171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,16384,0.031219555271996394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,12288,0.028070221344629925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,10240,0.025006221400366888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,8192,0.02088266611099243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,12288,0.02629511058330536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,7168,0.019163555569118924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,10240,0.02288533416059282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,6144,0.017269333203633625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,8192,0.02530844509601593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,5120,0.014998222390810648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,7168,0.02234755622016059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,4096,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,6144,0.02108977735042572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,3584,0.010631110933091907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,5120,0.019708444674809773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,3072,0.009407999614874521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,4096,0.01860000027550591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,3584,0.01700799994998508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,2560,0.00871022211180793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,2048,0.007663999994595845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,3072,0.015984000431166757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,1536,0.006360888895061281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,2560,0.014967999524540372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,1024,0.003960000144110786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,2048,0.014494222071435718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,768,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,1536,0.013516444298956128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,1024,0.012506666282812754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,512,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,768,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,512,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,256,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,128,0.002961777771512667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,256,0.011561777856614856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,64,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,128,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,6144,32,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,64,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,65536,0.107387555970086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,6144,32,0.011167111496130625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,51200,0.08846755822499593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,65536,0.07621510823567708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,16384,0.03484355409940084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,51200,0.061312887403700084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,12288,0.02606933315594991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,16384,0.02863911125395033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,10240,0.0221377776728736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,12288,0.023185777995321486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,10240,0.02176800039079454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,8192,0.019111111760139465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,7168,0.017257778180970084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,8192,0.023992000354660883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,6144,0.015214223000738354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,7168,0.021075555019908484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,5120,0.013908444179428948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,6144,0.019965334071053397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,4096,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,5120,0.01899377836121453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,4096,0.017251556118329365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,3584,0.009806222385830348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,3072,0.008756444685988957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,3584,0.01611111064751943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,2560,0.008249777886602614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,3072,0.014868444866604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,2048,0.007111999723646376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,2560,0.014663999279340109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,1536,0.0052968888647026485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,2048,0.013167111410035027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,1024,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,1536,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,768,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,1024,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,768,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,512,0.0032026666320032547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,256,0.002945777856641345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,512,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,128,0.003152888889114062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,256,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,64,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,128,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,5120,32,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,64,0.011139555937714048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,5120,32,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,65536,0.08722577492396037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,51200,0.06737422280841403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,65536,0.0695671108033922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,16384,0.027985778119828966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,51200,0.056355555852254234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,16384,0.025394666526052687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,12288,0.0207013338804245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,10240,0.0184897780418396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,12288,0.022016889519161646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,8192,0.015574221809705099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,10240,0.02019466625319587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,7168,0.013749333719412485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,8192,0.021760889225535925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,6144,0.01240355521440506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,7168,0.02050488856103685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,6144,0.018581334087583754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,5120,0.01129422254032559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,4096,0.009535111486911774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,5120,0.01796533332930671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,3584,0.008799110849698385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,4096,0.016523554921150208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,3072,0.008028444316652086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,3584,0.014647111296653748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,2560,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,3072,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,2560,0.013825777504179211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,2048,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,1536,0.004610666798220741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,2048,0.013182222015327878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,1024,0.00388711111413108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,1536,0.013181333740552267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,768,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,1024,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,512,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,768,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,256,0.003313777761326896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,512,0.01148799972401725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,128,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,256,0.011376000112957425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,128,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,64,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,4096,32,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,64,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,4096,32,0.011119999819331698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,65536,0.07640710804197523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,51200,0.0588595536020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,65536,0.06662933031717937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,16384,0.024393777052561443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,51200,0.05434044533305698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,12288,0.01854044364558326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,16384,0.024450666374630396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,10240,0.016390222642156813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,12288,0.02102488941616482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,8192,0.014162666267818876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,10240,0.019753777318530612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,7168,0.013767110804716745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,8192,0.022037333912319605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,7168,0.019183110859658983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,6144,0.01219555570019616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,5120,0.01075111081202825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,6144,0.017875555488798354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,4096,0.008987555901209513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,5120,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,3584,0.008432889150248634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,4096,0.014875556031862894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,3072,0.007652444144090016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,3584,0.014710222681363424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,3072,0.013648889131016202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,2560,0.006843555718660355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,2048,0.0053031109273433685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,2560,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,1536,0.004516444272465176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,2048,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,1024,0.003980444537268745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,1536,0.012513777448071374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,768,0.0035982223020659555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,1024,0.01183555523554484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,768,0.011388444238238864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,512,0.01145866678820716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,256,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,256,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,128,0.011132444772455426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,64,0.0025439999169773525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,64,0.011139555937714048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,512,0.0032755554550223877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3584,32,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,65536,0.062936888800727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3584,32,0.010954666468832227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,51200,0.05049688948525322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,65536,0.06429688798056708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,16384,0.021150221427281696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,51200,0.052463110950258046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,16384,0.023489778240521748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,12288,0.016593777471118502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,10240,0.01702577703528934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,12288,0.02064533366097344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,8192,0.014716444744004143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,8192,0.0200284438000785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,7168,0.012532444463835822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,7168,0.01861422260602315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,6144,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,6144,0.017475555340449016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,5120,0.010122666756312052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,5120,0.015863110621770222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,4096,0.008650666309727563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,4096,0.014528888795110913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,10240,0.01905422243807051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,3584,0.008243555823961893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,3072,0.0074515557951397365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,3584,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,3072,0.013474666410022311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,2560,0.006215999937719769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,2048,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,2560,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,1536,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,2048,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,1024,0.003967111309369405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,1536,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,768,0.003608889049953885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,1024,0.011593778100278644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,512,0.0031831111345026228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,768,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,256,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,512,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,256,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,128,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,64,0.0028320000403457214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,128,0.010760000182522668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,3072,32,0.002550222186578645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,64,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,65536,0.056141330136193164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,3072,32,0.010448888772063784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,51200,0.04602666695912679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,65536,0.061602665318383105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,51200,0.05036088824272156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,16384,0.019712888532214694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,12288,0.014940443966123792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,16384,0.023053333163261414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,10240,0.01385155568520228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,12288,0.02034311162100898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,8192,0.012156444291273752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,10240,0.018906666172875296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,8192,0.019521777828534443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,7168,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,7168,0.01797777745458815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,6144,0.010409778190983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,6144,0.016533333394262526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,5120,0.009563555320103964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,5120,0.014670222997665405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,4096,0.008408889174461365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,4096,0.013921777407328287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,3584,0.007687999970383114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,3584,0.013760888742076026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,3072,0.006953777538405524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,3072,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,2560,0.005653333332803514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,2560,0.013144888811641268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,2048,0.004913777940803104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,2048,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,1536,0.004253333227501975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,1536,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,1024,0.0038515557017591265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,768,0.0035075553589397003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,1024,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,512,0.0032595555401510666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,768,0.011495999991893768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,256,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,512,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,128,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,256,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,64,0.0026435556097163092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,128,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2560,32,0.0025519999778932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,32,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,65536,0.04595822095870972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,65536,0.05963910950554741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,51200,0.03814311159981622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,51200,0.04894666539298164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2560,64,0.011139555937714048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,16384,0.015398222539159985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,12288,0.012861332959598966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,16384,0.022912888063324824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,10240,0.013451555536852943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,12288,0.02032977839310964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,8192,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,10240,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,7168,0.011152889165613385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,8192,0.017940445078743827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,6144,0.010379555324713389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,7168,0.015561777684423657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,6144,0.0151555554734336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,5120,0.009731555150614845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,4096,0.0079644446571668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,5120,0.014508444401952954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,4096,0.013452444639470843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,3584,0.007270221908887227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,3072,0.006272000157170826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,3584,0.013482666677898832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,2560,0.005614222337802251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,3072,0.013175999952687157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,2048,0.0052177777720822235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,2560,0.012822222378518848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,1536,0.004388444539573458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,2048,0.012755555411179861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,1024,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,1536,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,768,0.003626666549179289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,1024,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,512,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,768,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,256,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,512,0.011872000164455838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,128,0.0030871110243929755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,64,0.0027031112048361036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,256,0.011112000379297467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,128,0.011800888511869641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,2048,32,0.0025537777692079544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,65536,0.03604622350798713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,64,0.011561777856614856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,2048,32,0.011340444286664328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,51200,0.029241777128643457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,16384,0.014866666661368476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,65536,0.057758225335015186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,51200,0.04760533240106371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,12288,0.01222222215599484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,16384,0.022468444373872545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,10240,0.011226666470368704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,8192,0.010032888915803697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,12288,0.019991111424234178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,10240,0.018301332990328472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,7168,0.010459555519951714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,8192,0.01591911084122128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,7168,0.015283554792404175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,5120,0.008728000024954477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,6144,0.01484444406297472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,4096,0.007311111523045435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,5120,0.014191110928853353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,3584,0.006728888799746831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,4096,0.01348533315791024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,3072,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,3584,0.01351111133893331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,2560,0.005606222069925732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,3072,0.012863111164834766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,6144,0.009690666364298927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,2048,0.0052986666560173035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,2560,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,1536,0.004597333156400257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,2048,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,1024,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,1536,0.012235555383894177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,768,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,1024,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,768,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,512,0.003261333331465721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,256,0.003463111196955045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,512,0.011122666299343109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,128,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,256,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,128,0.010786666638321347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1536,32,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,64,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,65536,0.025594666600227356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1536,32,0.010748444332016839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,51200,0.02156977852185567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,65536,0.05546311206287808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,16384,0.011500444677140979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,51200,0.04461510976155599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,12288,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,16384,0.02233422299226125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,10240,0.01034400032626258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,12288,0.018015111486117046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,8192,0.008988444175985124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,10240,0.01718933383623759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,7168,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,8192,0.015789333317014907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,6144,0.00869066682126787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,7168,0.015524443652894763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,5120,0.007722666694058313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,6144,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,4096,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,5120,0.014552889598740472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,3584,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,4096,0.013220444321632385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,3072,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,3584,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,2560,0.005594666633341048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,3072,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,2048,0.006317333214812809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,2560,0.012764444781674279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,1536,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,2048,0.012200000385443369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,1024,0.004311999926964442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,1536,0.012805333567990197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,768,0.004272888931963178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,1024,0.011123555401961008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,512,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,768,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,256,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,512,0.011110222174061669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,128,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,256,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,64,0.0027919999427265594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,128,0.010788444843557147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,1024,32,0.0029279999434947968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,64,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,65536,0.023426666855812073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,1024,32,0.010792888700962067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,51200,0.019575110740131803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,65536,0.05388444330957201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,16384,0.010588444769382477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,51200,0.044587555858823985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,12288,0.010382222632567087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,16384,0.021368889345063105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,10240,0.009761778016885122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,12288,0.018015111486117046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,8192,0.008384000096056197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,10240,0.017248888810475666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,7168,0.00870666652917862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,8192,0.015662221444977652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,6144,0.00831200016869439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,7168,0.015543111496501498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,5120,0.007655999726719326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,6144,0.014535999960369535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,4096,0.0069919998447100324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,3584,0.006654222392373615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,5120,0.014485332700941296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,3072,0.006015111174848344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,4096,0.013561777770519257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,3584,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,2560,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,2048,0.004919111314747068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,3072,0.013181333740552267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,2560,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,1536,0.004937777916590373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,2048,0.01205599970287747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,1024,0.0042568888101312844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,768,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,1024,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,1536,0.012201777762836881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,512,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,256,0.003056888985964987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,768,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,512,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,128,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,256,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,64,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,128,0.010772444307804108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,768,32,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,64,0.011472889118724398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,65536,0.01871466636657715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,768,32,0.010818666881985135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,51200,0.016543999314308167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,65536,0.053471999035941235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,16384,0.010023111270533668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,51200,0.04385155439376831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,12288,0.00998933364947637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,16384,0.020603555772039626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,10240,0.008772444393899705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,12288,0.01795733306143019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,8192,0.007898666792445714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,10240,0.01719288859102461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,7168,0.008767111433876885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,8192,0.01553422212600708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,6144,0.008236444658703273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,7168,0.01577333278126187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,5120,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,6144,0.014552889598740472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,4096,0.007278222176763747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,5120,0.014299554957283868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,3584,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,4096,0.01314755529165268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,3072,0.00629688882165485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,3584,0.01370044466522005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,2560,0.006301333506902059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,3072,0.012644444074895648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,2048,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,2560,0.012881778180599213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,1536,0.004267555558019214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,2048,0.011957333319716983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,1024,0.003690666622585721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,1536,0.012445333103338877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,768,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,1024,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,512,0.0033368888414568375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,768,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,256,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,512,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,128,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,256,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,64,0.0025519999778932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,128,0.010770666930410596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,512,32,0.0027475555737813315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,64,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,65536,0.016586666305859882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,512,32,0.010791999598344168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,51200,0.015503999259736804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,65536,0.05178755521774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,16384,0.00904799997806549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,51200,0.04080977704789903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,12288,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,16384,0.02067377832200792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,10240,0.007329777710967594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,12288,0.017883555756674874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,10240,0.01724177764521705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,8192,0.0063457778758472866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,7168,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,8192,0.015896000795894198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,6144,0.006999111009968652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,7168,0.015285332997639975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,5120,0.006331555545330048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,6144,0.014826666977670459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,4096,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,5120,0.013797333670987023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,3584,0.005727111051479976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,4096,0.013845333622561561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,3072,0.005902222047249476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,3584,0.013148444394270578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,3072,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,2560,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,2048,0.005345777918895085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,2560,0.012845333251688214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,1536,0.004543999830881755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,2048,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,1024,0.003919110943873723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,1536,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,768,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,1024,0.011178666518794166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,768,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,512,0.0118222220076455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,256,0.0029644444584846497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,256,0.01183644433816274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,128,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,64,0.0029484445436133277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,128,0.011702222128709158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,32,0.002546666603949335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,256,512,0.0032915555768542816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,64,0.01148000028398302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,65536,0.01421333352724711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,256,32,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,51200,0.012216889195972018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,16384,0.00793422261873881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,65536,0.04911377694871691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,51200,0.04046133491728041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,12288,0.007683555285135905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,16384,0.02071555621094174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,10240,0.007057777709431118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,12288,0.018283555905024212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,8192,0.006654222392373615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,10240,0.017073776986863878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,7168,0.006351111249791251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,8192,0.01589866644806332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,6144,0.006250666661394968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,7168,0.015208888385030957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,5120,0.005920889063013925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,6144,0.01479822231663598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,4096,0.0058959999846087555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,5120,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,3584,0.0058453331391016645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,4096,0.013516444298956128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,3072,0.0058906666106647914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,3584,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,2560,0.005941333456171884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,3072,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,2048,0.0058782220714622075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,2560,0.012788444757461548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,1536,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,2048,0.012488889197508493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,1024,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,1536,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,768,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,1024,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,512,0.0032755554550223877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,768,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,256,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,512,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,128,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,256,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,64,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,128,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,128,32,0.002586666701568498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,64,0.011375111010339526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,65536,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,128,32,0.011528888510333167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,51200,0.010092444717884064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,65536,0.048660443888770215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,16384,0.006807111203670502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,51200,0.04045333464940389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,12288,0.006999111009968652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,16384,0.020637333393096924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,10240,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,12288,0.018279999494552612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,8192,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,10240,0.017156443662113614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,7168,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,8192,0.015881778465376962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,6144,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,7168,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,5120,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,6144,0.014815110299322339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,4096,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,5120,0.014128888646761576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,3584,0.005759999983840519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,4096,0.013773333695199756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,3072,0.005904888941182031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,3584,0.013413333230548434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,2560,0.005964444329341252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,3072,0.013081777426931592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,2048,0.005647999958859549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,2560,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,1536,0.004256000121434529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,2048,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,1536,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,1024,0.0038355555799272326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,768,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,1024,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,512,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,768,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,256,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,512,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,128,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,256,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,64,0.002549333290921317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,128,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,64,32,0.002499555548032125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,64,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,65536,0.01070755554570092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,64,32,0.01073066641887029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,51200,0.010059555371602377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,65536,0.04866755670971341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,16384,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,51200,0.04014666544066535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,12288,0.006940444310506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,16384,0.020661332541041903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,10240,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,12288,0.018251554833518136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,8192,0.00664533343580034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,10240,0.01847555571132236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,7168,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,8192,0.015867556134859722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,6144,0.0063733334342638654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,7168,0.015459555718633862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,6144,0.014875556031862894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,5120,0.00591466658645206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,5120,0.014115555418862237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,4096,0.005570666657553778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,4096,0.013462222284740873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,3584,0.005579555614127054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,3584,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,3072,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,2560,0.0058782220714622075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,2560,0.012735111018021902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,3072,0.012669333153300814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,2048,0.005234666582610872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,1536,0.004249777644872665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,2048,0.011752000285519494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,1024,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,1536,0.01180622229973475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,768,0.003510222252872255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,1024,0.010991111397743225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,512,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,768,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,256,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,512,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,256,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,128,0.0028044444819291434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,64,0.002455111179086897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,15,32,32,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,64,0.010427555276287926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,15,32,32,0.011033777561452655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,51200,0.49605512619018555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,51200,0.9526293012830945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,16384,0.3366568883260091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,12288,0.2262995507982042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,16384,0.16798933347066244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,10240,0.1895893282360501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,12288,0.1304746601316664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,8192,0.15260622236463758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,10240,0.11189866065979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,7168,0.13410133785671657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,7168,0.10097422202428182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,8192,0.1300871107313368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,6144,0.11572444438934326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,6144,0.09541777769724528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,5120,0.09810400009155273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,4096,0.07804355356428358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,5120,0.0837751097149319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,3584,0.06933155324723986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,4096,0.06487555636299981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,3072,0.06015377574496799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,3584,0.05561955769856771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,2560,0.05089066757096184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,3072,0.053130666414896645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,2048,0.041625777880350746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,2560,0.04825955629348755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,1536,0.03257600135273404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,2048,0.041965332296159535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,1024,0.023131555981106226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,1536,0.03301066822475857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,768,0.018553778529167175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,1024,0.025831111603313025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,512,0.013236444857385425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,768,0.021728888154029846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,256,0.008912888665994009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,512,0.018978666928079393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,128,0.006270222365856171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,256,0.015966221690177917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,64,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,128,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,65536,32,0.005921777751710679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,64,0.015122666954994202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,65536,32,0.015558222929636637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,65536,0.5231377813551161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,65536,1.0317796071370442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,51200,0.8009591102600098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,51200,0.4018809000651042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,16384,0.2757199870215522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,16384,0.13751199510362414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,12288,0.20143911573621961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,12288,0.10725244548585679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,10240,0.16731733745998809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,8192,0.13675288359324136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,10240,0.0923466682434082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,7168,0.1173057821061876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,8192,0.10607644584443833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,7168,0.08291021982828777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,6144,0.1032222244474623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,5120,0.08805511395136516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,6144,0.07856977648205228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,4096,0.0711031092537774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,5120,0.07061955663892958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,4096,0.05794577466117012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,3584,0.06241689125696818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,3072,0.052793777651256986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,3584,0.048047112094031445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,2560,0.044646223386128746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,2048,0.03705244594150119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,2560,0.041617777612474226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,1536,0.029442667961120605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,2048,0.03335466649797227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,1024,0.020816889074113634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,3072,0.04586044285032484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,1536,0.029721776644388836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,768,0.01646488904953003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,768,0.02012711101108127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,512,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,1024,0.022657778528001573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,512,0.017296888762050204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,256,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,128,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,128,0.012793777717484368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,64,0.0036346668170558084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,64,0.013199999928474426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,32,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,51200,32,0.0144213338692983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,51200,256,0.00756888919406467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,65536,0.3312266667683919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,51200,0.2602337731255425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,65536,0.1833111047744751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,16384,0.09695288870069717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,51200,0.1403119961420695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,12288,0.07018044259813097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,16384,0.054810666375690036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,12288,0.04478844337993198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,10240,0.061075554953681103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,10240,0.03970400161213345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,8192,0.04317422376738655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,7168,0.037490665912628174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,8192,0.04410488737954033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,6144,0.03296355406443278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,5120,0.028432889117134943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,7168,0.039346665143966675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,6144,0.0355662206808726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,4096,0.02327555583582984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,5120,0.03211999932924906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,3584,0.02109066645304362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,3072,0.018024888303544786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,3584,0.024233778317769367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,4096,0.027479999595218237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,2560,0.0158906661801868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,2048,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,2560,0.021644444929228887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,3072,0.023152889476882085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,2048,0.019345778557989333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,1536,0.010741333166758219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,1024,0.008352888955010308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,768,0.0069395556218094295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,1536,0.017242666747834947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,1024,0.015377778146002026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,768,0.014027555783589682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,512,0.004936888813972473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,256,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,128,0.0031795555518733132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,256,0.012000888586044312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,128,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,64,0.002971555623743269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,512,0.013289777768982781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,16384,32,0.0032275555034478507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,64,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,16384,32,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,65536,0.24678667386372885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,65536,0.15654667218526205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,51200,0.19391822814941406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,16384,0.07050844695832995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,12288,0.05137066708670723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,51200,0.12196621629926895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,16384,0.04878222280078464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,10240,0.04384266667895847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,12288,0.03998311029540168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,8192,0.0372284452120463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,7168,0.03276000089115567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,10240,0.036144001616372004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,8192,0.03976266582806905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,6144,0.028997331857681274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,5120,0.024101333485709295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,6144,0.03111999895837572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,7168,0.03427022364404466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,4096,0.019855999284320407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,3584,0.017518222332000732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,5120,0.030039999220106337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,3072,0.015574221809705099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,4096,0.025158221522967022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,3584,0.021848888860808477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,2560,0.013767110804716745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,2048,0.011560888753996955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,3072,0.021635555558734473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,2560,0.019811555743217468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,1536,0.009734222458468543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,1024,0.007725333174069722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,2048,0.017351999878883362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,1536,0.0165928883685006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,768,0.006620444357395172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,1024,0.014442667365074158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,512,0.004300444490379757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,256,0.003809777812825309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,768,0.013208889298968844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,128,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,512,0.012755555411179861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,128,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,256,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,32,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,64,0.011122666299343109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,12288,32,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,65536,0.20723556147681343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,65536,0.14815110630459255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,12288,64,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,51200,0.16137866179148355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,16384,0.06366222434573703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,51200,0.11741777261098225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,16384,0.047084444099002414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,12288,0.04603111081653171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,10240,0.0388995541466607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,12288,0.03852355480194092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,10240,0.035016000270843506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,8192,0.03162044286727905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,7168,0.029337777031792536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,8192,0.03324800067477756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,6144,0.02622577713595496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,7168,0.02898755669593811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,5120,0.022802667485343084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,5120,0.024290665984153748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,6144,0.026995556222067937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,4096,0.0221377776728736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,3584,0.016198222835858662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,3584,0.020741333564122517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,3072,0.014434667097197639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,3072,0.019676445258988273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,2560,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,4096,0.01812622282240126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,2048,0.011220444407727985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,2560,0.018186666899257235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,1536,0.00920800036854214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,2048,0.016504888733228047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,1024,0.007471111085679796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,1536,0.015598222613334656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,768,0.005393777870469623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,1024,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,512,0.004826666580306159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,768,0.013171555267439948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,512,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,256,0.0038008888562520346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,256,0.011805333197116852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,64,0.0029768887907266617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,128,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,10240,32,0.0031537777847713898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,64,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,65536,0.16347289085388184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,10240,32,0.01204622205760744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,51200,0.1292817725075616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,65536,0.10121333599090576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,16384,0.05106222298410204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,51200,0.0790897806485494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,12288,0.032607999112870954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,16384,0.03501777847607931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,10240,0.028355555401908025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,12288,0.029926220575968426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,10240,0.026905778381559584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,8192,0.023361777265866596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,7168,0.02088000045882331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,8192,0.026169778572188482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,6144,0.018546667363908555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,6144,0.02274311085542043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,7168,0.024619556135601465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,5120,0.015927111109097798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,4096,0.01312711089849472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,5120,0.02236622240808275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,4096,0.018572444717089336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,3584,0.011939555406570435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,3072,0.01093777765830358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,3584,0.017904000149832833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,2560,0.00962755580743154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,3072,0.01754577789041731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,2048,0.008430221842394935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,2560,0.016240888171725802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,1536,0.007048000064161088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,2048,0.015559110376569959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,1024,0.0052488889131281115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,1024,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,768,0.003944000022278892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,768,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,512,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,1536,0.013516444298956128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,256,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,512,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,128,0.003177777760558658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,256,0.011197333534558615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,64,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,8192,32,0.0029493332323100832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,32,0.01184266640080346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,65536,0.15541956159803602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,65536,0.09294044309192234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,51200,0.1150782240761651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,128,0.011809777882364062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,16384,0.042873776621288724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,51200,0.0728693339559767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,12288,0.03327644533581204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,16384,0.03330399923854404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,12288,0.028029332558314007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,10240,0.02887822190920512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,8192,0.024089778463045757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,10240,0.025103110406133864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,8192,0.025417778227064345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,7168,0.02169688873820835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,8192,64,0.010695111420419482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,6144,0.016906667086813185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,7168,0.02400355537732442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,5120,0.014737778239780001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,6144,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,4096,0.0122417774465349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,5120,0.02069155540731218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,3584,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,4096,0.018603555030292936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,3072,0.010235555469989777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,3584,0.01724088854259915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,2560,0.00920800036854214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,3072,0.01718488832314809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,2048,0.008045333127180735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,2560,0.016574222180578444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,1536,0.006752000086837345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,2048,0.014159111513031853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,1024,0.004411555412742826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,1536,0.013556444810496436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,768,0.0038844446341196695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,1024,0.01239111108912362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,512,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,768,0.012089778151777057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,256,0.003485333174467087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,512,0.011828444070286222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,128,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,256,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,64,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,128,0.01074399964676963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,7168,32,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,64,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,7168,32,0.010683555569913654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,65536,0.12881510787540013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,51200,0.0949279997083876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,65536,0.08368444442749023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,16384,0.03661333190070258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,51200,0.06640977991951837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,16384,0.030926220946841772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,12288,0.028340445624457464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,10240,0.02461244497034285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,12288,0.026195555925369263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,8192,0.020992888344658747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,10240,0.02274666726589203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,7168,0.018722666634453666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,8192,0.025455999705526564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,6144,0.017200888858901132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,7168,0.02275111112329695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,5120,0.01514488955338796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,6144,0.021037333541446265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,4096,0.01129333343770769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,5120,0.02016533414522807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,3584,0.010658666491508484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,4096,0.018040888839297824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,3072,0.009483555952707926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,3584,0.016858667135238647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,2560,0.00869955536392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,3072,0.01666933298110962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,2048,0.007621333003044128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,2560,0.015470221638679504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,1536,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,2048,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,1024,0.0041857775714662345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,1536,0.013071999781661563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,768,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,1024,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,512,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,768,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,256,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,512,0.011657777759763928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,128,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,256,0.011344888971911537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,64,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,128,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,6144,32,0.0028035555862718155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,64,0.011376889215575324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,6144,32,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,65536,0.10673689179950291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,51200,0.08751910924911499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,65536,0.07586933506859674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,16384,0.03470666540993585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,51200,0.061158219973246254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,12288,0.025810667210155066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,16384,0.028879112667507593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,10240,0.022145777940750122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,12288,0.023394667439990576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,8192,0.019119110372331407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,10240,0.021362667282422382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,7168,0.017265778448846605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,8192,0.02371911042266422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,6144,0.015408888459205627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,7168,0.021670222282409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,5120,0.013890667094124688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,6144,0.020807999703619216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,4096,0.010495999621020423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,5120,0.019174221489164565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,3584,0.00983199973901113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,4096,0.01741155485312144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,3584,0.016143111719025504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,3072,0.008735111190213097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,3072,0.015118221441904703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,2560,0.008047111332416534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,2560,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,2048,0.0070568886068132185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,2048,0.013429332938459186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,1536,0.0052968888647026485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,1536,0.013008000122176277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,1024,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,1024,0.012059555285506778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,768,0.0037582222786214617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,768,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,512,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,512,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,256,0.0032506665835777917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,256,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,128,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,128,0.010791111323568555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,64,0.00254844439526399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,64,0.010382222632567087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,5120,32,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,5120,32,0.010500444306267632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,65536,0.08671733405854966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,65536,0.06899466779496935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,51200,0.06708444489373101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,51200,0.05620888868967692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,16384,0.0288080010149214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,16384,0.026103110777007207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,12288,0.021220443977249995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,12288,0.02166666587193807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,10240,0.018586667047606576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,10240,0.02036177780893114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,8192,0.015628443823920358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,8192,0.02221244408024682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,7168,0.013745778136783175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,7168,0.020279111133681405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,6144,0.012430222498046027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,6144,0.018965333700180054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,5120,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,5120,0.01819022165404426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,4096,0.009355555805895064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,4096,0.01592266725169288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,3584,0.008788444101810455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,3584,0.014814222852389017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,3072,0.008080888953473832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,3072,0.013781333135233985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,2560,0.007277333074145847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,2560,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,2048,0.006210666563775804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,1536,0.0048071112897661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,2048,0.012854222622182635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,1024,0.004212444441186057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,1024,0.011775111158688864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,768,0.0038604442444112566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,1536,0.013258666627936892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,512,0.0034435554924938413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,768,0.01145866678820716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,256,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,256,0.011153777440388998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,128,0.002954666606254048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,512,0.011947555674446953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,64,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,128,0.010784889260927836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,4096,32,0.0028968888024489083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,32,0.010471111370457543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,4096,64,0.011939555406570435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,65536,0.07633599970075819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,51200,0.05882844660017225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,51200,0.05403022302521599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,65536,0.06646933158238728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,16384,0.024890666206677754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,12288,0.01867466668287913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,12288,0.021334222621387903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,10240,0.016349333855840895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,16384,0.02476444509294298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,8192,0.014325333966149224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,10240,0.019653333557976615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,7168,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,8192,0.022086222966512043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,7168,0.01958755486541324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,6144,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,5120,0.010633777413103314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,6144,0.018366222580273945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,4096,0.009038222332795462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,5120,0.0181342214345932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,4096,0.01500888831085629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,3584,0.008355555435021719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,3072,0.007679111427730984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,3584,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,2560,0.006875555548402999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,3072,0.014259555273585849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,2560,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,2048,0.005642666584915585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,1536,0.004635555462704765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,2048,0.012725333372751871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,1024,0.0038933331767717996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,1536,0.01312800000111262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,768,0.00369422220521503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,768,0.012243555651770698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,1024,0.011735111474990845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,512,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,512,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,256,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,128,0.010760889285140567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,64,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,64,0.010382222632567087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3584,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3584,32,0.010676444404655032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,65536,0.06304177973005506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,65536,0.06364177995257907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,51200,0.05061511198679606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,51200,0.05174577898449368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,16384,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,16384,0.023576888773176406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,12288,0.016937777400016785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,12288,0.020703110429975722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,10240,0.017295999659432303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,10240,0.019331556227472093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,8192,0.014873777826627096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,8192,0.020447999238967896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,7168,0.012847999731699625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,7168,0.01862844493654039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,6144,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,6144,0.018002667360835604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,5120,0.010320888625250923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,5120,0.016752888758977253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,4096,0.00870400004916721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,4096,0.014581332604090372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,3584,0.008227555288208855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,3584,0.013557333085272046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,3072,0.007140444384680853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,2560,0.006113777971929974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,3072,0.013454222016864352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,2048,0.005437333136796951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,2560,0.013178666432698568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,2048,0.01255733354224099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,1536,0.004595555365085602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,1024,0.003960888832807541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,1536,0.012320000264379712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,768,0.0035457776652442086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,1024,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,512,0.003257777748836411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,768,0.011367999845080905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,256,0.0031306667046414483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,512,0.0114702218108707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,128,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,256,0.011104889214038849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,64,0.0025271111064487034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,128,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,3072,32,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,64,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,65536,0.05607022179497612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,3072,32,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,51200,0.045824001232783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,65536,0.06106311082839966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,16384,0.01971733404530419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,51200,0.05037600133154127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,12288,0.014764444695578681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,16384,0.023074666659037273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,10240,0.013842666314707862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,12288,0.02050133380624983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,10240,0.018799111247062683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,8192,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,7168,0.011349333657158745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,8192,0.019039111004935372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,6144,0.0103493332862854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,5120,0.00977955593003167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,6144,0.016887111796273124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,4096,0.008210666477680206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,5120,0.01536799967288971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,3584,0.007695111135641734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,4096,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,3072,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,3584,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,2560,0.005960888746711943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,3072,0.013198222551080914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,2048,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,2560,0.012248888611793518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,1536,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,2048,0.013015111287434896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,1536,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,1024,0.00388355553150177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,768,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,1024,0.011620444556077322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,512,0.003495111233658261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,768,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,256,0.0029537777105967202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,512,0.01110577748881446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,128,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,256,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,64,0.002537777854336633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,7168,0.01779822177357144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,128,0.011004444625642566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2560,32,0.003104000041882197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,64,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2560,32,0.010320888625250923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,65536,0.04633955491913689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,51200,0.03801600138346354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,65536,0.0591848889986674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,16384,0.014814222852389017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,51200,0.04832444588343302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,12288,0.01275199982855055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,16384,0.022675555613305833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,10240,0.013334222137928009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,12288,0.019920888874265883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,8192,0.011620444556077322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,10240,0.018823110395007663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,7168,0.011047999891969891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,8192,0.017889777819315594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,6144,0.010258666343159145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,7168,0.01591999994383918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,5120,0.009523555636405945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,6144,0.014825777875052558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,4096,0.008184889124499427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,5120,0.014196443888876172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,3584,0.006992888947327931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,4096,0.013456000222100152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,3072,0.00629688882165485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,3584,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,2560,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,3072,0.012826667063766055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,2048,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,2560,0.012724444270133972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,1536,0.004579555657174853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,2048,0.012067555553383298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,1024,0.003839111162556542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,1536,0.012061333490742577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,768,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,1024,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,512,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,768,0.011704000333944956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,256,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,512,0.011150221857759686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,128,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,256,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,64,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,128,0.01071466671095954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,2048,32,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,64,0.011143999795118967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,65536,0.036215109957589045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,2048,32,0.010421333213647207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,51200,0.029425776667065088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,65536,0.057447108957502574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,16384,0.01457333399189843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,51200,0.04686133397950066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,12288,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,16384,0.022648000054889258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,10240,0.011052444577217102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,12288,0.019730667273203533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,8192,0.010383110907342698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,10240,0.018181333939234417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,7168,0.010194666683673859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,8192,0.016186666157510545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,6144,0.009375111096435124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,7168,0.015449777245521545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,5120,0.00870666652917862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,6144,0.014518222875065274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,4096,0.007339555356237624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,5120,0.014104889498816596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,3584,0.006598222172922558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,4096,0.013448889056841532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,3072,0.006294222341643439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,3584,0.01312888910373052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,2560,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,3072,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,2048,0.005179555465777715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,2560,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,1536,0.0042408886882993905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,2048,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,1024,0.0038444445365005066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,1536,0.0120666664507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,768,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,1024,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,512,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,768,0.01167555567291048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,256,0.0031351111829280853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,512,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,128,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,256,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,64,0.0027591110103660156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,128,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1536,32,0.0028044444819291434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,64,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,65536,0.025563556287023757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,65536,0.05505422088834974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,51200,0.021783999270863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,51200,0.044172445933024086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,16384,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,16384,0.02164799968401591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,12288,0.011035555766688453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,12288,0.01827822294500139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,10240,0.010206222534179688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1536,32,0.010535111029942831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,8192,0.008721777962313758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,10240,0.01682400041156345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,7168,0.009296000003814697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,8192,0.015621332658661736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,7168,0.015279110934999255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,6144,0.008569777839713627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,5120,0.007999999655617608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,6144,0.014851555228233337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,4096,0.007293333609898885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,4096,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,3584,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,5120,0.014504889647165934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,3584,0.013080000049538083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,3072,0.0064444442590077715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,2560,0.005771555420425203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,3072,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,2048,0.0064035554726918536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,2560,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,2048,0.012431110772821637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,1536,0.006078222145636876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,1024,0.0042275554604000514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,1536,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,768,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,1024,0.011824888487656912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,512,0.0035075553589397003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,768,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,256,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,512,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,256,0.011377777490350934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,128,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,64,0.003097777772280905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,128,0.011503111157152386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,1024,32,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,64,0.011704888608720569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,1024,32,0.010757333702511258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,65536,0.02392533255947961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,51200,0.01919911139541202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,65536,0.05355377660857307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,16384,0.010538666612572141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,51200,0.044235554006364614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,12288,0.01052977806992001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,16384,0.020827554994159274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,10240,0.00923733330435223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,12288,0.018236445056067575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,10240,0.017127111554145813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,8192,0.00833422193924586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,7168,0.008721777962313758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,8192,0.01586577792962392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,6144,0.008010666403505538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,7168,0.01571377780702379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,5120,0.007666666474607255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,6144,0.014762666490342883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,4096,0.0069751110341813825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,5120,0.014427555931939019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,3584,0.006633777585294511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,4096,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,3072,0.00600355573826366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,3584,0.0135057775510682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,3072,0.013171555267439948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,2560,0.005773333211739858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,2048,0.005224888937340842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,2560,0.012527999778588613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,1536,0.004931555440028508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,2048,0.012401777837011548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,1024,0.004239999999602635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,1536,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,768,0.003903999924659729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,1024,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,512,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,768,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,256,0.0032355555643637977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,512,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,256,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,128,0.0031431110368834603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,64,0.0026684444811609057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,128,0.011518222590287527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,768,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,64,0.011608888705571493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,65536,0.018735999862353008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,768,32,0.011069333387745751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,51200,0.016915554801623028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,65536,0.05313511027230156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,51200,0.043819556633631386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,16384,0.020439111524158053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,12288,0.009727111293209923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,12288,0.018239110708236694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,10240,0.008965333302815756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,10240,0.016883555385801528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,8192,0.007974222302436829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,8192,0.01575555569595761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,16384,0.010118222071064843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,7168,0.00834222220712238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,6144,0.00833333366447025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,7168,0.015244444211324057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,5120,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,6144,0.014658666319317289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,4096,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,5120,0.014134221606784396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,4096,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,3584,0.0069475554757648045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,3072,0.0063288890653186375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,3584,0.013072888884279462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,2560,0.005973333285914526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,3072,0.013060444758998023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,2560,0.012741333080662621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,2048,0.005744888964626525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,2048,0.012073777616024017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,1536,0.004259555704063839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,1536,0.01181688904762268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,1024,0.0036720000207424164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,1024,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,768,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,512,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,512,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,256,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,768,0.011239999698268043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,128,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,256,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,64,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,64,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,128,0.010807111031479306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,512,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,65536,0.016566221912701923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,65536,0.049787554475996226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,51200,0.015522667103343539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,51200,0.04076977901988559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,16384,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,16384,0.020326221982638042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,12288,0.008055999875068665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,12288,0.018273777431911893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,10240,0.007344000041484833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,512,32,0.010060444474220276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,10240,0.017095110482639737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,8192,0.006321777900060018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,7168,0.006256000035338932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,8192,0.015879111157523263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,6144,0.006976889239417181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,7168,0.015238222148683337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,6144,0.014561777313550314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,5120,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,4096,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,5120,0.014138667119873894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,3584,0.00599644457300504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,4096,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,3584,0.013097777962684631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,3072,0.005510222166776657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,2560,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,3072,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,2048,0.005655999812814925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,2560,0.011904000408119626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,1536,0.004250666747490565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,1536,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,1024,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,1024,0.011217777927716574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,768,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,768,0.0111191107167138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,2048,0.012496888637542725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,512,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,512,0.011271110839313932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,128,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,256,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,64,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,128,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,256,32,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,32,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,256,64,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,65536,0.014145778285132514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,51200,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,51200,0.04049688908788893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,65536,0.04872177706824409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,16384,0.007986666427718269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,12288,0.007675555845101674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,12288,0.01824711097611321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,16384,0.01997155613369412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,10240,0.007286222444640265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,8192,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,8192,0.015824889143308003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,7168,0.006284444282452266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,10240,0.016695110334290397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,6144,0.0058373332851462895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,7168,0.015222221612930298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,5120,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,6144,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,4096,0.005274666680230035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,5120,0.014456889695591398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,3584,0.005927111125654644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,4096,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,3584,0.01313244468635983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,3072,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,2560,0.005601777798599667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,3072,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,2048,0.005721777677536011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,2560,0.012314666476514606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,1536,0.00423911131090588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,2048,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,1024,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,1536,0.011744000017642975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,768,0.003301333429084884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,1024,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,512,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,256,0.003007111036115222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,768,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,512,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,128,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,256,0.01089244418674045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,64,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,128,0.01070755554570092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,128,32,0.002558222247494592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,64,0.010348444183667501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,65536,0.010942222343550788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,128,32,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,51200,0.010093332992659675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,65536,0.047671112749311656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,16384,0.006732444382376141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,51200,0.039834668238957725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,16384,0.020026667250527274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,12288,0.006666666517655055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,10240,0.006249777972698212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,12288,0.017975999249352347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,8192,0.005976888868543837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,10240,0.016619554824299283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,7168,0.005643555687533484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,8192,0.015562666787041558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,6144,0.005671999934646819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,7168,0.01518133282661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,5120,0.005394666559166378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,6144,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,4096,0.0052684446175893145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,5120,0.013831111292044321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,3584,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,4096,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,3072,0.005221333354711533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,3584,0.01308444473478529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,2560,0.005986666513813867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,3072,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,2048,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,2560,0.01257155587275823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,1536,0.004289777742491828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,2048,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,1024,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,1536,0.01183733344078064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,1024,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,768,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,512,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,768,0.011074666347768573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,256,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,512,0.010491555763615502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,128,0.002546666603949335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,256,0.011023999916182624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,64,0.0025537777692079544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,128,0.010760889285140567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,64,32,0.002534222271707323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,64,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,65536,0.010807111031479306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,64,32,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,51200,0.010094222095277574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,65536,0.04764444298214383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,16384,0.006834666762087081
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,51200,0.03981333308749729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,16384,0.019966221517986722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,12288,0.006960888703664144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,10240,0.006276444428496891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,12288,0.017808000246683758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,8192,0.006249777972698212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,10240,0.016602666841612924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,7168,0.005602666487296422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,8192,0.015503111812803479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,6144,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,7168,0.015164444843928019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,6144,0.014183110660976835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,5120,0.005315555466545953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,4096,0.005446222093370225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,5120,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,3584,0.005594666633341048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,4096,0.013261333107948303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,3072,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,3072,0.012866666747464074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,2560,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,2560,0.012423999607563019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,2048,0.005305777821275923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,2048,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,1536,0.004305777864323722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,3584,0.012856889102194043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,1024,0.003555555724435382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,1536,0.01186222251918581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,768,0.003573333223660787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,1024,0.011021333436171213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,512,0.0030088888274298776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,768,0.011168888873524137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,512,0.010685332947307162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,256,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,128,0.0026480000880029467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,256,0.010755555497275459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,64,0.002534222271707323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,128,0.010780444575680627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,14,32,32,0.002562666725781229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,64,0.010455999937322406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,14,32,32,0.010871111518806882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,51200,0.4959511227077908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,16384,0.33770399623446995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,51200,0.9525395499335395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,12288,0.22616267204284668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,16384,0.16820622815026176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,12288,0.13031644291347927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,10240,0.18948354985978869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,10240,0.1118897729449802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,8192,0.1522622240914239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,7168,0.13416622744666204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,8192,0.11991555160946316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,6144,0.11582844787173802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,7168,0.10132266415490045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,6144,0.090865777598487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,5120,0.09817422098583645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,4096,0.07804355356428358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,5120,0.08243466748131646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,3584,0.06924000051286486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,4096,0.06773955292171903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,3584,0.059697780344221324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,3072,0.05571022298600939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,2560,0.05111111203829447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,2560,0.043551110559039645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,2048,0.041802667909198336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,2048,0.04120088948143853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,1536,0.03243911266326904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,3072,0.059980445437961154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,1024,0.023120888405376013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,1536,0.034295111894607544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,768,0.018388445178667705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,1024,0.025272000167104933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,512,0.013436444103717804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,768,0.020661332541041903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,256,0.008802666432327693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,256,0.01536977787812551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,128,0.006199999815887875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,512,0.018872888551818002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,64,0.00564533347884814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,128,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,65536,32,0.005630222045713001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,64,0.015559110376569959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,65536,32,0.015169777803950839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,65536,0.523746649424235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,65536,1.0279217825995552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,51200,0.40177422099643284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,51200,0.7997271219889323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,16384,0.27241333325703937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,12288,0.19711111651526558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,16384,0.1372684505250719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,10240,0.16675110658009848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,12288,0.10695733626683553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,10240,0.09210666682985093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,8192,0.1350791056950887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,7168,0.11731199423472087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,7168,0.08241866694556342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,6144,0.10189688867992824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,8192,0.09867288668950398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,6144,0.07912800047132704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,5120,0.08618666728337605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,4096,0.07116355498631795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,3584,0.06225422355863783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,5120,0.07304444577958849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,3072,0.052912887599733144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,4096,0.05396177702479892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,3584,0.049865778949525624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,2560,0.04498222139146593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,2048,0.037125332487954035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,3072,0.04581244455443489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,1536,0.0288355549176534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,2560,0.03729866610633002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,1024,0.02186222208870782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,2048,0.035176889763938055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,1536,0.029334220621320937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,1024,0.022427555587556627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,512,0.011853333148691388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,768,0.019183999962276883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,256,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,512,0.01715377800994449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,128,0.003931555483076307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,256,0.01478666729397244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,64,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,128,0.012866666747464074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,32,0.0035537779331207275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,64,0.014860444598727755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,51200,32,0.014308444327778287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,51200,768,0.01607377827167511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,65536,0.33064267370435924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,65536,0.18324444029066297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,51200,0.2602880001068115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,16384,0.09584444761276245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,12288,0.07045866383446588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,51200,0.14082845052083334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,16384,0.054568890068266124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,10240,0.06028266747792562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,8192,0.042972445487976074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,10240,0.04007733199331496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,7168,0.038020445240868464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,8192,0.04522488845719231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,6144,0.03308711118168301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,7168,0.039114667309655085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,5120,0.028652442826165095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,6144,0.036880890528361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,4096,0.02332266668478648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,5120,0.03223289052645365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,3584,0.020981333321995203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,4096,0.02774133284886678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,12288,0.04459555612670051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,3072,0.018230222993426852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,3584,0.025313778056038752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,3072,0.024262222978803847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,2560,0.015615999698638916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,2048,0.013206221991115145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,2560,0.020761777957280476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,1536,0.010911111202504901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,2048,0.019849777221679688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,1536,0.017603556315104168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,1024,0.014848889576064216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,768,0.007032000356250339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,768,0.013291555146376291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,512,0.004888888862397936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,512,0.012852444416946836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,256,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,256,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,1024,0.008433777425024245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,128,0.003237333355678452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,64,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,64,0.01108444482088089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,128,0.012058667010731168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,16384,32,0.0031431110368834603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,16384,32,0.012440000143316058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,65536,0.24697687890794542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,51200,0.194304002655877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,65536,0.15843110614352757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,16384,0.06906577613618639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,51200,0.12097956074608697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,16384,0.04849511053827074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,12288,0.0510844455824958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,10240,0.0440684457619985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,12288,0.04020444552103678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,10240,0.035543110635545515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,8192,0.03722577624850803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,7168,0.03286222285694546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,8192,0.03808800048298306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,6144,0.029088000456492107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,7168,0.03369066781467862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,5120,0.02380888826317257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,6144,0.030116445488399927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,4096,0.019651555352740817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,5120,0.028949333561791316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,3584,0.017622222503026325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,4096,0.025161777933438618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,3072,0.01552888916598426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,3584,0.022989332675933838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,3072,0.022202667262819078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,2560,0.013653332988421122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,2048,0.011706666813956367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,2560,0.019094222121768527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,1536,0.009729777773221334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,2048,0.018526222970750596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,1024,0.007879111501905654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,1536,0.016392888294325936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,768,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,1024,0.014463999205165438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,512,0.004567999806669024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,768,0.013059555656380124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,256,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,512,0.012827555338541666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,256,0.01147822207874722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,128,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,64,0.0031679999083280563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,128,0.011062222222487131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,12288,32,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,64,0.012378666963842181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,12288,32,0.011111111276679568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,65536,0.2059555583530002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,65536,0.1473217805226644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,51200,0.16083378261990017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,16384,0.06440977917777167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,51200,0.11706488662295872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,12288,0.045760889848073326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,16384,0.047188444270028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,10240,0.03871555460823907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,12288,0.038804445001814104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,8192,0.031537777847713895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,10240,0.03494400117132399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,8192,0.0328053335348765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,7168,0.029061334000693426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,6144,0.02603200078010559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,7168,0.028838220569822524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,5120,0.022847111026446026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,6144,0.027257778578334387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,4096,0.018016000588734944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,5120,0.02445777753988902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,3584,0.01624000072479248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,4096,0.022346667117542688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,3584,0.020959110723601446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,3072,0.0144533332851198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,2560,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,3072,0.019796444310082328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,2048,0.011008888483047485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,2560,0.018914666440751817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,1536,0.009298666483826106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,2048,0.016187555260128446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,1024,0.0075235557225015424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,1024,0.013016000390052795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,1536,0.01533511115445031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,768,0.005904888941182031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,768,0.013178666432698568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,512,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,512,0.012806221842765808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,256,0.0038693332009845306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,256,0.012226666841242047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,128,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,128,0.01180355581972334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,64,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,64,0.012429333395428128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,10240,32,0.002951111023624738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,10240,32,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,65536,0.16299200057983398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,65536,0.10109421941969131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,51200,0.1289235618379381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,51200,0.07899822129143609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,16384,0.0512951115767161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,16384,0.035346666971842446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,12288,0.032391998502943255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,12288,0.029512001408471003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,10240,0.027990221977233887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,10240,0.026914666096369427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,8192,0.023401778605249193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,8192,0.026141333911154006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,7168,0.02091466718249851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,7168,0.02446311049991184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,6144,0.01832711034350925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,6144,0.022882666852739122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,5120,0.015967110792795818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,5120,0.02160888910293579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,4096,0.013364444176355997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,4096,0.019525332583321463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,3584,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,3584,0.01868355605337355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,3072,0.010997333460383944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,3072,0.01761866609255473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,2560,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,2560,0.016935999194780987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,2048,0.008634666601816813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,2048,0.015524443652894763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,1536,0.007317333585686154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,1536,0.01387288918097814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,1024,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,768,0.004062222109900581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,1024,0.012813333008024426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,512,0.003617777592606015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,768,0.01200799975130293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,512,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,256,0.0032720000793536506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,128,0.002999111182159848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,256,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,64,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,128,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,8192,32,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,64,0.011815110842386881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,8192,32,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,65536,0.15271466308169895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,65536,0.09210755427678426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,51200,0.11487289269765218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,16384,0.042619556188583374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,51200,0.0724453330039978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,12288,0.033246222469541765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,16384,0.03309422069125705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,10240,0.028785778416527644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,12288,0.028027554353078205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,8192,0.02411288850837284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,10240,0.024300444457266066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,7168,0.021895110607147217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,8192,0.02714666724205017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,6144,0.01680266691578759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,7168,0.024433778391944036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,5120,0.014895111322402954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,6144,0.022758222288555566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,4096,0.012263110942310758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,5120,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,3584,0.011300444602966309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,4096,0.018616000811258953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,3072,0.010272888673676385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,3584,0.018163555198245578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,2560,0.009230222139093611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,3072,0.017623111605644226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,2048,0.007978666987684038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,2560,0.016506666938463848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,1536,0.006894222150246303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,2048,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,1024,0.004276444514592488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,1536,0.013849777479966482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,768,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,1024,0.012772444221708508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,512,0.003240888938307762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,768,0.012624888784355588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,512,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,256,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,256,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,64,0.0025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,128,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,7168,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,64,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,7168,32,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,65536,0.1291982200410631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,51200,0.09520710839165582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,65536,0.08412444591522217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,16384,0.03675288955370585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,51200,0.06660799847708808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,12288,0.028210666444566514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,16384,0.030950221750471327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,10240,0.024446222517225478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,12288,0.025264889001846313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,8192,0.021017778250906203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,10240,0.023090667194790308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,7168,0.01896711190541585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,8192,0.025157334076033697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,6144,0.01683466633160909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,7168,0.023031999667485554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,5120,0.014975999792416891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,6144,0.022083555658658344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,4096,0.011352889239788055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,5120,0.020011555817392137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,3584,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,4096,0.018602665927675035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,3072,0.009450666606426239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,3584,0.017913776967260573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,2560,0.008591111335489485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,3072,0.016919111212094624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,2048,0.007481777833567725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,2560,0.015899555550681222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,2048,0.014554666148291694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,1536,0.006638222270541721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,1024,0.003955555458863576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,1536,0.013453333742088742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,1024,0.0121742222044203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,768,0.003671110918124517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,512,0.003391999958289994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,768,0.011858666936556498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,512,0.012142221960756512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,256,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,128,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,256,0.011715555356608497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,64,0.0028000000036425064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,128,0.010823999842007955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,6144,32,0.0026168889469570587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,64,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,6144,32,0.011183111204041375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,65536,0.10696444246504043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,51200,0.08729866478178237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,65536,0.07583555248048571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,16384,0.03503377901183234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,51200,0.06100533405939738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,12288,0.026000888811217412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,16384,0.028456889920764502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,10240,0.02204533252451155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,12288,0.023220444718996685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,8192,0.019090667366981506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,10240,0.021586666504542034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,7168,0.017330666383107502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,8192,0.023711111810472276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,6144,0.015503999259736804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,7168,0.021576889687114294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,6144,0.020937777227825586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,5120,0.013838222457302941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,4096,0.010553778045707278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,3584,0.009891555541091496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,5120,0.019256888164414298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,4096,0.017214222086800467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,3072,0.00870488915178511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,2560,0.00814488861295912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,3584,0.016527111331621807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,2048,0.007279111279381647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,3072,0.015879111157523263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,2560,0.013794666363133324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,1536,0.005050666630268097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,2048,0.01384622189733717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,1024,0.003923555629120933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,768,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,1536,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,512,0.0033804445217053094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,1024,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,768,0.011818666425016193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,256,0.0029502221279674103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,512,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,128,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,64,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,256,0.011128000087208219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,128,0.010763555765151978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,5120,32,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,64,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,65536,0.08640088637669881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,5120,32,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,51200,0.06722222434149848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,16384,0.028414222929212783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,65536,0.06982666916317411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,51200,0.05597155623965793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,12288,0.02088000045882331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,10240,0.018612444400787354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,16384,0.025409777959187824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,12288,0.021798221601380244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,8192,0.01552888916598426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,7168,0.013982221484184265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,10240,0.020657777786254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,8192,0.021720889541837905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,6144,0.012490666574902005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,7168,0.020229332976871066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,5120,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,6144,0.018906666172875296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,4096,0.009376000199053023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,5120,0.01836266616980235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,4096,0.016689777374267578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,3584,0.008671999805503422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,3072,0.008092444803979661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,3584,0.01479555500878228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,2560,0.007425777614116669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,3072,0.014845333165592618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,2048,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,2560,0.013424889081054263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,1536,0.004549333204825719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,2048,0.013581333061059317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,1024,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,1536,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,768,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,1024,0.012441777520709567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,768,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,512,0.0032551110618644287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,256,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,512,0.011384888655609555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,128,0.002806222273243798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,256,0.011722666521867117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,64,0.0028035555862718155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,128,0.010802666346232096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,4096,32,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,64,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,4096,32,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,65536,0.0762231085035536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,51200,0.05874311261706882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,65536,0.06568888823191325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,16384,0.024360888534122046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,51200,0.05401955710517036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,12288,0.018592000007629395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,16384,0.024187554915746052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,10240,0.016158221496476066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,12288,0.02144533395767212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,8192,0.013986666997273764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,10240,0.019555555449591745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,7168,0.013823111024167804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,8192,0.021939555803934734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,7168,0.019536889261669584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,6144,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,5120,0.01068800025516086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,6144,0.018277333842383493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,4096,0.008948444492287105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,5120,0.018019555343521964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,3584,0.00851111114025116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,4096,0.01567555632856157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,3072,0.007827555967701806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,3584,0.014734221829308404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,2560,0.006959999601046245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,3072,0.013767110804716745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,2048,0.005614222337802251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,2560,0.014177777700954013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,1536,0.004549333204825719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,2048,0.012658666405412885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,1024,0.004110222061475118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,1536,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,1024,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,768,0.003718222181002299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,512,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,768,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,256,0.002973333415057924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,512,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,128,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,256,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,64,0.0025600000388092464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,128,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3584,32,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,64,0.010414222048388587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,65536,0.06270488765504625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3584,32,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,51200,0.05049688948525322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,65536,0.06371111339992948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,16384,0.021300445000330608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,51200,0.05209066801600986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,12288,0.01645511057641771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,16384,0.02379200028048621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,10240,0.017251556118329365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,12288,0.02097333305411869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,10240,0.019281778070661757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,8192,0.014630221658282809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,7168,0.012450666891203986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,8192,0.01998488936159346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,6144,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,7168,0.018668444620238412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,5120,0.010117332968446944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,6144,0.017453332742055256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,4096,0.008866666919655269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,5120,0.016137777103318107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,3584,0.008280889027648503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,4096,0.015127110812399121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,3072,0.007381333245171442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,3584,0.014105778601434497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,3072,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,2560,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,2560,0.013156444662147097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,2048,0.004933333231343163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,2048,0.012490666574902005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,1536,0.004544888933499654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,1536,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,1024,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,1024,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,768,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,768,0.011809777882364062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,512,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,512,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,256,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,256,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,128,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,128,0.010715555813577441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,64,0.0031191110610961914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,3072,32,0.002704888789190186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,64,0.010447999669445885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,65536,0.05591111050711738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,3072,32,0.010818666881985135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,51200,0.04570400052600437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,65536,0.06066133578618368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,16384,0.019631110959582858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,51200,0.050070222881105214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,12288,0.015169777803950839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,16384,0.023215999205907185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,10240,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,12288,0.020469332734743755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,8192,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,10240,0.018970666660202872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,7168,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,8192,0.018954666124449838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,6144,0.010482666393121084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,7168,0.017628444565667045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,5120,0.009883555273214975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,6144,0.016208888755904306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,4096,0.008362666600280339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,5120,0.0151573336786694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,3584,0.00796888851457172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,4096,0.013843555417325763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,3072,0.006672888994216919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,3584,0.013468444347381592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,2560,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,3072,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,2048,0.0052177777720822235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,2560,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,1536,0.004536889079544279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,2048,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,1024,0.003943110919660992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,1536,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,768,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,1024,0.012048000262843238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,512,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,768,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,256,0.0031235555393828284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,512,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,128,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,256,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,64,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,128,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2560,32,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,64,0.010404444403118556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,65536,0.046201778782738574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2560,32,0.010808000134097205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,51200,0.03791822327507867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,65536,0.05874400006400215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,16384,0.015075556106037565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,51200,0.04842311143875122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,12288,0.012674666941165924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,16384,0.0226275556617313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,10240,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,12288,0.02036355601416694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,8192,0.011552000211344825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,10240,0.018556444181336295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,7168,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,8192,0.01759288873937395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,6144,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,7168,0.01651822196112739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,5120,0.009372444616423713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,6144,0.01482488877243466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,4096,0.007676444119877285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,5120,0.014521777629852295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,3584,0.007140444384680853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,4096,0.013606222139464485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,3072,0.006019555446174409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,3584,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,2560,0.005531555662552516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,3072,0.01313511116637124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,2048,0.004919111314747068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,2560,0.012840000291665396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,1536,0.004495999879307217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,2048,0.012079999678664737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,1024,0.0038648889296584656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,1536,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,768,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,1024,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,512,0.0031831111345026228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,768,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,512,0.011687999798191918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,256,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,256,0.01108888867828581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,128,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,128,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,64,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,64,0.011167111496130625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,2048,32,0.002493333278430833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,2048,32,0.011021333436171213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,65536,0.03568622138765123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,65536,0.057256890667809375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,51200,0.029331554969151814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,51200,0.04691822330156962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,16384,0.014878221684032016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,16384,0.022312889496485393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,12288,0.01218311074707243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,12288,0.019890666007995605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,10240,0.011334222224023608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,10240,0.018206222189797294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,8192,0.010115555591053432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,8192,0.01590044465329912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,7168,0.010354667074150508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,7168,0.015199111567603217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,6144,0.009687999884287516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,6144,0.014814222852389017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,5120,0.008310221963458592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,5120,0.014175999495718213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,4096,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,4096,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,3584,0.006646222124497096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,3584,0.013445333474212222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,3072,0.005910222315125995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,3072,0.013099555340078143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,2560,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,2048,0.00517155561182234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,2560,0.012806221842765808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,1536,0.004447111239035924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,1536,0.012087999946541257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,1024,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,2048,0.011983999775515662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,1024,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,768,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,512,0.0032666667054096856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,768,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,256,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,512,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,128,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,256,0.011069333387745751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,64,0.00254488881263468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,128,0.010822222464614444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,64,0.011085333095656501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1536,32,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,65536,0.02546666728125678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1536,32,0.010777778095669217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,51200,0.021446221404605444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,65536,0.05485066771507263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,16384,0.011174221833546957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,51200,0.044215109613206655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,12288,0.010840000377760993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,16384,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,12288,0.018352889352374606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,10240,0.00999466660949919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,8192,0.00905688852071762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,10240,0.016857778032620747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,7168,0.009371555513805812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,8192,0.01567555632856157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,6144,0.008426666259765625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,7168,0.015141333142916361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,5120,0.007628444168302748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,6144,0.014688889185587565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,4096,0.007336888876226213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,4096,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,3584,0.006589333216349284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,3584,0.013209777573744455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,3072,0.005952888892756567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,3072,0.012771555946932899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,2560,0.005576888720194499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,5120,0.014310222533014087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,2048,0.006263999889294307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,2560,0.012806221842765808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,1536,0.005481777919663324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,2048,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,1024,0.004280000097221798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,1536,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,768,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,1024,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,512,0.003620444486538569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,768,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,256,0.0033039999090962936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,512,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,128,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,256,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,64,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,64,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,1024,32,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,32,0.011174221833546957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,65536,0.023768888579474554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,1024,128,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,51200,0.019484443797005545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,65536,0.05364533265431722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,51200,0.043884442912207715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,16384,0.010619555910428366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,12288,0.010618666807810465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,12288,0.01847022275129954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,10240,0.009535999761687385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,16384,0.02102222210831112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,8192,0.008360889222886827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,10240,0.017364444004164804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,7168,0.008904000123341879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,8192,0.01587022178702884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,6144,0.008004444340864817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,6144,0.014504000544548035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,5120,0.007375111182530721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,7168,0.01564800077014499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,4096,0.007155555817815993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,5120,0.014503111441930136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,3584,0.00665155549844106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,3584,0.013526221944226159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,4096,0.013882666826248169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,3072,0.0062871111763848206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,2560,0.005569777968857024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,2560,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,2048,0.00490044429898262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,3072,0.013187555803192986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,1536,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,2048,0.011753777662913004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,1024,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,1024,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,768,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,1536,0.012220444778601328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,512,0.003549333247873518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,768,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,256,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,256,0.010805333654085795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,128,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,512,0.011872889267073737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,64,0.002883555574549569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,128,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,768,32,0.0029351111087534162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,32,0.011049778097205691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,768,64,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,65536,0.01889955500761668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,51200,0.016809778081046212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,51200,0.043567111094792686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,16384,0.009927111367384592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,65536,0.0529440012243059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,12288,0.009883555273214975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,16384,0.020311999652120803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,10240,0.009022222624884712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,10240,0.016872889465755887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,8192,0.008036444584528605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,12288,0.01833511061138577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,7168,0.00906399968597624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,8192,0.015861334072219003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,6144,0.00811111099190182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,6144,0.014660444524553088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,7168,0.015607999430762397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,5120,0.007401777638329401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,4096,0.00702400008837382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,3584,0.006697777658700943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,4096,0.013352889153692456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,5120,0.014381332529915703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,3072,0.006633777585294511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,2560,0.006280000011126201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,3072,0.013188444077968597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,2048,0.005960000058015187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,2048,0.012543111211723752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,1536,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,1536,0.012306667036480375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,3584,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,1024,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,768,0.003510222252872255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,1024,0.011734222372372946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,512,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,2560,0.01237155579858356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,256,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,768,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,256,0.010420444111029306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,64,0.0025519999778932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,64,0.010051555931568146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,512,32,0.0026088888860411118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,128,0.011647111011876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,65536,0.016569778323173523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,32,0.010068444742096795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,51200,0.015499555402331881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,51200,0.03983199927541945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,16384,0.00869511150651508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,65536,0.051501333713531494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,12288,0.007803555164072249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,512,512,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,10240,0.007308444215191736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,16384,0.019896000623703003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,8192,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,10240,0.01685599982738495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,8192,0.015535111228624979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,7168,0.006619555668698416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,6144,0.006647999915811751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,7168,0.015122666954994202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,6144,0.014448889427714877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,12288,0.01828977796766493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,5120,0.006270222365856171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,5120,0.01439377831088172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,4096,0.005679111099905438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,3584,0.005986666513813867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,3584,0.012768000364303589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,3072,0.005609777652555042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,4096,0.013845333622561561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,2560,0.006464889066086874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,3072,0.012461333639091916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,2048,0.005531555662552516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,2048,0.011691555380821228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,2560,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,1536,0.004508444418509801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,1024,0.003924444317817688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,1024,0.011015111373530494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,1536,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,768,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,512,0.0032542221662071017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,512,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,256,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,768,0.012278222375445895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,128,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,256,0.010392888552612728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,64,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,128,0.011530666715568967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,64,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,256,32,0.002785777880085839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,65536,0.014376888672510782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,256,32,0.010123555858929953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,51200,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,51200,0.039317333035998873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,65536,0.04867911007669237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,16384,0.007655999726719326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,12288,0.007373332977294922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,12288,0.017639999588330586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,10240,0.0070266665683852295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,10240,0.016548444827397663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,8192,0.006393777827421824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,8192,0.0155395550860299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,7168,0.006001777946949005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,7168,0.014918223023414612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,6144,0.005943111247486538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,16384,0.0206977774699529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,6144,0.014448000325096978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,5120,0.0063164445261160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,4096,0.005769777629110549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,5120,0.01348888874053955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,3584,0.005684444473849402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,3584,0.012813333008024426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,4096,0.013604444762070974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,3072,0.005682666682534748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,2560,0.005624889085690181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,3072,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,2048,0.005706666658322017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,2048,0.012572444147533841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,1536,0.004442666553788715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,1024,0.004040889028045866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,1536,0.012159110771285163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,768,0.0035048888789282907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,1024,0.011184888581434885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,512,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,768,0.012173333101802401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,256,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,2560,0.012415111064910889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,128,0.002528888897763358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,512,0.011111111276679568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,64,0.002557333351837264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,256,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,128,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,128,32,0.002559111143151919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,65536,0.011014222270912595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,64,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,128,32,0.010060444474220276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,51200,0.01018044435315662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,16384,0.006629333313968446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,65536,0.048278222481409706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,51200,0.03911022345225016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,12288,0.007322666545708974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,10240,0.006693333387374878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,16384,0.020015999674797058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,12288,0.017610667480362784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,8192,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,7168,0.005942222144868638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,10240,0.01718133356836107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,8192,0.015473778049151102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,7168,0.014888000157144336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,5120,0.0053119998839166425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,6144,0.014167999227841696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,4096,0.005254222287072076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,5120,0.013818666338920593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,4096,0.013088000317414602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,3584,0.005630222045713001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,3072,0.005593777944644292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,3584,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,2560,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,3072,0.012500444220172035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,2560,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,2048,0.0053004444473319584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,1536,0.004305777864323722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,6144,0.006248000181383557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,2048,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,1024,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,1536,0.01218488895230823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,768,0.003547555456558863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,768,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,512,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,512,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,256,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,256,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,1024,0.011026666396194033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,128,0.010442666709423065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,64,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,64,0.01039644413524204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,32,0.0026106666773557663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,64,32,0.010067555639478896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,65536,0.0107342220014996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,64,128,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,65536,0.0476746658484141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,51200,0.010092444717884064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,16384,0.00665866666369968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,51200,0.039092444711261325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,12288,0.0063573333124319715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,16384,0.02067733307679494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,10240,0.006993778049945831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,12288,0.01759288873937395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,8192,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,10240,0.01722577710946401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,8192,0.015568888849682279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,7168,0.007040888898902469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,6144,0.006314666734801398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,7168,0.014911111858155994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,5120,0.0052871112194326185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,6144,0.014870221416155497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,4096,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,5120,0.013535111314720578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,3584,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,4096,0.013812444276279874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,3584,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,3072,0.005622222191757626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,2560,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,3072,0.012470222181744046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,2048,0.005315555466545953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,2560,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,1536,0.004549333204825719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,2048,0.01146044416560067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,1024,0.0037600000699361167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,1536,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,1024,0.011042666931947073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,768,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,512,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,768,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,256,0.0027795556104845474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,512,0.011622221933470832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,128,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,256,0.010458666417333813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,64,0.00253866674999396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,128,0.011492444409264458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,64,0.010415111151006486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,13,32,32,0.002752888947725296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,13,32,32,0.010012444522645738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,51200,0.4950915442572699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,51200,0.9503733317057291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,16384,0.33606844478183323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,16384,0.16780266496870253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,12288,0.22593598895602754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,12288,0.12998933262295193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,16384,0.18726489279005262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,10240,0.1938880019717746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,12288,0.14760800202687582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,10240,0.17338044113583037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,8192,0.15250221888224283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,8192,0.11882400512695312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,10240,0.13023555278778076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,7168,0.13382755385504827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,51200,0.5165475739373101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,8192,0.10762400097317165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,7168,0.10196977853775024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,6144,0.11573066976335313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,6144,0.10506844520568848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,5120,0.09800977839363946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,7168,0.09730666875839233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,6144,0.08759199910693699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,5120,0.08016266425450642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,4096,0.07789955536524455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,4096,0.06491644514931573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,3584,0.06891822152667575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,5120,0.07735821935865614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,3584,0.06085599793328179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,4096,0.0670693318049113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,3072,0.05984977881113688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,3584,0.06140622165468004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,3072,0.05499200026194254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,2560,0.05073688758744133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,3072,0.05715911255942451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,2048,0.04148088892300924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,2560,0.04819555415047539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,2560,0.05082044336530897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,1536,0.03224710954560174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,2048,0.03719822234577603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,1536,0.0325164430671268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,2048,0.04702044526735941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,1024,0.02296177711751726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,1536,0.040292445156309344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,1024,0.02296177711751726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,768,0.018312000566058688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,1024,0.03336266676584879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,768,0.022643556197484333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,768,0.030984888474146526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,512,0.013236444857385425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,512,0.029355555772781372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,256,0.00870311094654931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,512,0.018838221828142803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,256,0.025065778030289546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,256,0.01621955633163452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,128,0.006276444428496891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,65536,128,0.020056888461112976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,64,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,128,0.014184888866212634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,65536,32,0.005617777920431561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,64,0.014522666732470194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,65536,32,0.01623911162217458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,65536,0.5236266454060873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,65536,1.0279369354248047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,51200,0.8010569148593478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,51200,0.40053155687120223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,65536,0.5271120071411133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,16384,0.272307554880778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,51200,0.40723466873168945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,16384,0.13737066586812338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,12288,0.19773244857788086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,12288,0.1071279976103041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,16384,0.1509635580910577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,10240,0.16771911250220406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,10240,0.10584888855616252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,8192,0.13668978214263916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,12288,0.12055822213490804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,8192,0.11072444253497654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,7168,0.11852888266245525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,10240,0.10528000195821126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,7168,0.08597866694132487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,8192,0.08959111240175034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,6144,0.10424533155229355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,6144,0.08292533291710748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,5120,0.08541066779030694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,7168,0.08216710885365804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,6144,0.07352533605363634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,5120,0.07265688975652058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,4096,0.07272266679339938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,4096,0.056464003192053906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,5120,0.06638489166895549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,3584,0.061193777455223926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,4096,0.05753066804673937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,3584,0.0512951115767161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,3072,0.05483822358979119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,3584,0.05504177676306831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,3072,0.04888000090916952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,2560,0.04492444462246365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,3072,0.04965600040223864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,2048,0.03862133291032579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,2560,0.041032890478769936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,2560,0.04788711004787021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,2048,0.032590223683251276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,1536,0.02871555421087477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,1536,0.028024888700909082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,1024,0.021454221672481958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,1536,0.03867022196451823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,1024,0.022292445103327434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,768,0.016505777835845947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,1024,0.03253511256641812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,768,0.01973155637582143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,512,0.011861333416567909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,768,0.030360887447992962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,512,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,256,0.007655999726719326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,2048,0.042007111840777926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,512,0.028384000062942505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,128,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,256,0.013436444103717804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,256,0.022693332698610093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,64,0.003536888708670934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,128,0.013409777647919126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,51200,128,0.01996977792845832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,51200,32,0.0036071112586392295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,32,0.014102222190962898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,51200,64,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,65536,0.3308497799767388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,65536,0.18272977405124238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,51200,0.2596408791012234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,51200,0.14017600483364528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,16384,0.09661689069535996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,16384,0.05436444282531738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,16384,0.06754310925801595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,12288,0.07033689154518975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,65536,0.193648894627889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,12288,0.04480266571044922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,51200,0.1527715524037679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,10240,0.06034666962093777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,8192,0.04322310951020983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,10240,0.05696088737911648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,12288,0.05767377879884508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,10240,0.052721778551737465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,7168,0.03776977790726556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,8192,0.04371910956170824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,8192,0.04726755619049072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,7168,0.03937600056330363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,6144,0.033026665449142456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,7168,0.04569244384765625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,6144,0.03526399864090814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,5120,0.028671112325456407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,6144,0.042428443829218544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,4096,0.023374223046832617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,5120,0.030945777893066406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,5120,0.039304888910717435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,3584,0.021054221524132624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,4096,0.034825778669781156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,4096,0.026536888546413843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,3072,0.018146667215559218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,3584,0.026007110873858135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,3584,0.033039109574423894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,2560,0.015771556231710646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,3072,0.023954666323131982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,3072,0.03263911273744371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,2048,0.012844444149070315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,2560,0.021631111701329548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,2560,0.030642665094799463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,1536,0.010344888601038191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,2048,0.028297778632905748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,2048,0.018567999203999836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,1024,0.008273777862389883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,1536,0.017495999733606975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,1024,0.013975110318925647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,768,0.007143999967310164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,1024,0.02556177808178796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,768,0.01350488927629259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,768,0.024021334118313257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,512,0.004938666605287128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,512,0.02204533252451155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,256,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,512,0.012630222572220696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,256,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,128,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,256,0.021189333664046392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,128,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,1536,0.027056889401541814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,16384,128,0.021179555190934077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,64,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,16384,32,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,64,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,16384,32,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,65536,0.2458639939626058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,65536,0.15607466962602404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,51200,0.19418666097852919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,51200,0.12125066916147868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,16384,0.07045066356658936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,16384,0.048723555273479886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,16384,0.057890666855706104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,65536,0.15746222601996526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,12288,0.05079822076691521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,51200,0.12362933158874512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,12288,0.03995022177696228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,12288,0.04920444554752774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,10240,0.04718133476045397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,8192,0.03689333465364244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,10240,0.047019554509056934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,8192,0.037608888414171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,7168,0.03252355588807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,8192,0.04225511021084256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,7168,0.03460444344414605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,6144,0.028581334484948054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,7168,0.040222220950656466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,6144,0.03248622351222568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,10240,0.04407111141416761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,5120,0.024147555232048035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,5120,0.027669333749347266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,5120,0.0355057782597012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,4096,0.019652444455358718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,4096,0.024271999796231587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,3584,0.017581333716710407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,4096,0.03241599930657281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,3584,0.02370577719476488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,3072,0.01553244392077128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,3072,0.021920889616012573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,3584,0.032102220588260226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,6144,0.0369395547442966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,2560,0.01332266628742218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,3072,0.03089244498146905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,2048,0.011671110987663269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,2560,0.01995822290579478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,2560,0.029373334513770208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,1536,0.009389333426952362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,2048,0.02741777731312646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,1536,0.026605332891146343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,1536,0.016588444511095684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,1024,0.007683555285135905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,1024,0.013077333569526672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,1024,0.022016000416543748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,768,0.0064693333374129395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,768,0.01294311136007309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,768,0.021983999345037673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,2048,0.017285333739386666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,512,0.012576888832781049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,512,0.021687999367713928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,256,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,256,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,512,0.005227555417352253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,128,0.0035075553589397003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,256,0.021846221552954778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,128,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,64,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,12288,128,0.020281778441535104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,12288,32,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,64,0.011029333704047732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,12288,32,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,65536,0.20484799808926055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,65536,0.14998932679494223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,51200,0.1604186693827311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,51200,0.11667111184861924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,65536,0.13953333430820042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,16384,0.06366399923960368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,16384,0.046839998828040234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,51200,0.10982666412989299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,12288,0.045774221420288086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,16384,0.053519109884897866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,12288,0.038690666357676186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,10240,0.03864799936612447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,12288,0.046395556794272534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,10240,0.039055999782350324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,8192,0.0314995547135671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,10240,0.04416977696948581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,8192,0.03228622343805101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,7168,0.02908088763554891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,8192,0.039064887497160174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,7168,0.029528889391157363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,7168,0.037725332710478045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,6144,0.026181333594852026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,6144,0.02631644407908122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,6144,0.035423109928766884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,5120,0.02269422180122799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,5120,0.02403555479314592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,5120,0.033005333609051175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,4096,0.017910222212473553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,4096,0.02174577779240078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,4096,0.03186577889654372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,3584,0.021041777398851182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,3584,0.03100088900989956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,3072,0.014030221435758801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,3072,0.019670221540662978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,2560,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,2560,0.018060444129837882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,3584,0.01603822244538201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,2048,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,2560,0.028831998507181805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,3072,0.030349334081013996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,1536,0.009014222357008193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,2048,0.016327111257447135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,2048,0.02788533435927497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,1024,0.007357333269384172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,1536,0.026127111580636766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,1536,0.015164444843928019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,1024,0.012988444831636218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,768,0.006250666661394968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,1024,0.021360889077186584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,512,0.0047475554876857335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,768,0.012529777983824411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,768,0.023066666391160753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,512,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,256,0.0038977778620190094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,512,0.021714667479197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,128,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,256,0.010591111249393888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,256,0.02197155521975623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,64,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,128,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,10240,128,0.018613333503405254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,10240,32,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,32,0.01035199976629681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,65536,0.16324622101253933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,10240,64,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,51200,0.12870666715833876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,65536,0.10097689098782009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,51200,0.07878933350245158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,16384,0.050556444459491305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,16384,0.0351475543446011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,65536,0.1241733365588718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,51200,0.10143733024597168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,12288,0.032255109813478254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,12288,0.029559999704360962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,10240,0.028083556228213843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,16384,0.05160977774196201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,12288,0.044269333283106484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,8192,0.023032888770103455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,10240,0.031056000126732722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,10240,0.04191111193762886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,8192,0.02628622286849552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,7168,0.020841777324676514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,8192,0.03820177912712097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,6144,0.018574222922325134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,7168,0.024086222052574158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,7168,0.0366542206870185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,6144,0.022479999396536086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,6144,0.03462666604253981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,5120,0.016231111354298063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,5120,0.03296000096533034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,5120,0.020643555455737643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,4096,0.01292888902955585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,4096,0.018926221463415358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,4096,0.03070755468474494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,3584,0.011907555162906647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,3584,0.018182223041852314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,3584,0.030509332815806072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,3072,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,3072,0.017467555072572496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,3072,0.029513776302337646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,2560,0.009487110707494948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,2560,0.016404444972674053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,2560,0.028824888997607764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,2048,0.008323555191357931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,2048,0.01443199978934394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,1536,0.007241778075695038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,1536,0.013771555489963956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,1536,0.025639111797014873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,1024,0.005578666511509154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,1024,0.0124177775449223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,1024,0.022078222698635522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,768,0.0041786668201287585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,2048,0.026509332987997267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,768,0.012192000117566852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,512,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,768,0.02104533380932278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,512,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,512,0.019905777441130746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,256,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,256,0.010421333213647207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,256,0.021309332715140447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,128,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,128,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,8192,128,0.01856355534659492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,64,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,8192,32,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,64,0.010694222317801582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,8192,32,0.010690666735172272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,65536,0.15135733286539713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,65536,0.09126310878329807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,51200,0.11428711149427627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,51200,0.0721928874651591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,65536,0.12175467279222275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,16384,0.04248088929388258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,16384,0.03278755479388767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,51200,0.10107733143700494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,12288,0.03313422203063965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,16384,0.049695998430252075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,12288,0.028055111567179363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,10240,0.028868443436092798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,12288,0.043235556946860425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,10240,0.03044711218939887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,8192,0.024241778585645888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,10240,0.04143822193145752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,8192,0.02534666657447815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,8192,0.03700800074471368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,7168,0.02167555524243249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,7168,0.023704888092146978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,7168,0.03674844569630093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,6144,0.01699555582470364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,6144,0.022991999983787537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,6144,0.033654222885767617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,5120,0.014727110664049784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,5120,0.02036888897418976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,5120,0.03317866722742716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,4096,0.012406222522258759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,4096,0.01832355558872223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,4096,0.03058755397796631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,3584,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,3584,0.017826666434605915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,3584,0.030187557140986126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,3072,0.010024000373151567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,3072,0.02902311086654663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,3072,0.01719466679626041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,2560,0.008986666798591614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,2560,0.016200888488027785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,2560,0.027385777897304956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,2048,0.007989333735571967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,2048,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,2048,0.02712533374627431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,1536,0.006892444358931647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,1536,0.013540444274743399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,1024,0.00526577772365676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,1024,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,1024,0.021983999345037673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,768,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,768,0.01205244412024816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,768,0.02201333310869005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,512,0.0035297779573334586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,1536,0.023557333482636347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,512,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,256,0.003173333282272021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,512,0.020290666156344943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,256,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,256,0.019945777124828763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,128,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,128,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,7168,128,0.018975110517607797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,64,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,64,0.010670222342014313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,7168,32,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,7168,32,0.010098666780524785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,65536,0.12872533003489176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,65536,0.08266755607393053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,51200,0.09576800134446885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,51200,0.0657857788933648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,65536,0.12148711416456436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,16384,0.036962668100992836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,16384,0.030977778964572485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,51200,0.10007288720872666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,12288,0.028042667441897925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,16384,0.04932622114817301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,12288,0.025919111238585577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,12288,0.04284711016549004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,10240,0.02492622203297085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,10240,0.027515555421511333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,8192,0.020982222424613103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,10240,0.039868444204330444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,8192,0.025495110286606684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,8192,0.036794665786955096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,7168,0.022720888257026672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,7168,0.03552533189455668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,6144,0.01663466625743442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,6144,0.02237333357334137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,7168,0.019108444452285767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,5120,0.014896000425020853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,6144,0.034146666526794434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,5120,0.020022221737437777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,4096,0.011266666981909009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,4096,0.018269333574506972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,5120,0.03270399901602004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,3584,0.010136889086829292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,4096,0.03048622277047899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,3584,0.029543111721674602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,3584,0.017608000172509086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,3072,0.0090506664580769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,3072,0.02882755464977688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,2560,0.008399999803966945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,3072,0.016864889197879367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,2560,0.015826667348543804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,2048,0.007506666911972894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,2560,0.02815999918513828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,2048,0.013502221968438892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,1536,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,2048,0.02555022140343984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,1536,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,1024,0.004303110970391167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,1536,0.023489778240521748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,1024,0.011975111232863532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,768,0.004005333201752769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,768,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,1024,0.021983999345037673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,512,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,768,0.02102222210831112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,512,0.020036444067955017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,256,0.0033208889265855155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,512,0.011477332976129321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,256,0.010407110883129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,128,0.002979555477698644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,256,0.020643555455737643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,128,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,6144,128,0.020416888925764296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,6144,32,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,64,0.010784000158309937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,65536,0.10660888751347859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,6144,32,0.010956444674068026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,65536,0.07527911000781588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,51200,0.08779999944898818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,51200,0.06009066767162747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,65536,0.12094577153523763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,16384,0.03477422065205044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,51200,0.09982933600743611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,16384,0.02889066603448656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,12288,0.025703110628657874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,16384,0.04925866590605842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,12288,0.04216088851292928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,12288,0.0236151119073232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,10240,0.022092445029152766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,10240,0.04039644532733493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,8192,0.019161777363883126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,10240,0.027582221561008032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,8192,0.02436622149414486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,7168,0.01721955504682329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,8192,0.03646399908595615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,7168,0.02180266711446974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,7168,0.035191999541388616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,6144,0.015400888191329109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,6144,0.03304355674319797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,5120,0.014001778430408902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,6144,0.021087111698256597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,5120,0.019115555617544387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,4096,0.010598222414652506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,5120,0.03229066729545593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,4096,0.017539555827776592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,4096,0.030264887544843886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,3584,0.009940444595283931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,3584,0.016530666086408827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,3584,0.029174221886528864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,3072,0.00903555585278405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,3072,0.015178667174445258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,3072,0.02834933333926731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,2560,0.008392000363932716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,2560,0.014528888795110913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,2560,0.02548711167441474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,2048,0.006701333241330252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,2048,0.013376889129479727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,2048,0.023369777533743117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,1536,0.005286222116814719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,1536,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,1536,0.023071999351183575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,1024,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,1024,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,1024,0.020782222350438435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,768,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,768,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,768,0.02107911143038008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,512,0.0032968889507982465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,512,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,512,0.0199608885579639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,256,0.0031928889867332247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,256,0.01035288886891471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,256,0.01998044384850396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,128,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,128,0.010911111202504901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,5120,128,0.018516444497638278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,64,0.002785777880085839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,64,0.010440000229411654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,5120,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,5120,32,0.010042666561073726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,65536,0.08637244171566433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,65536,0.06862666871812609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,51200,0.06671644581688775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,51200,0.05580800109439426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,65536,0.12084710597991943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,16384,0.028318223026063707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,16384,0.025136889682875738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,51200,0.09975288973914252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,12288,0.021038222644064162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,16384,0.0488319993019104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,12288,0.021831111775504217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,10240,0.018326222896575928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,12288,0.042316443390316434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,10240,0.024013333850436743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,8192,0.015303111738628812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,10240,0.04038933250639174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,8192,0.022587555978033278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,8192,0.03769244419203864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,7168,0.01385155568520228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,7168,0.02030311193731096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,7168,0.03487733337614272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,6144,0.01290933373901579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,6144,0.018923555811246235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,6144,0.03349333339267307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,5120,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,5120,0.031722668144438006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,5120,0.01763288842307197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,4096,0.009357333183288574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,4096,0.016487111647923786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,3584,0.008954666554927826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,3584,0.014841778410805596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,3584,0.02929333183500502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,3072,0.007996444072988298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,3072,0.01479822231663598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,3072,0.027120888233184814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,2560,0.007069333559936947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,4096,0.029919998513327703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,2560,0.013859555953078799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,2048,0.006176888942718506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,2560,0.02473955518669552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,2048,0.0127368892232577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,2048,0.023349333140585158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,1536,0.004886222382386525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,1536,0.01203644441233741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,1536,0.02331022255950504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,1024,0.003931555483076307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,1024,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,1024,0.020942222740915086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,768,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,768,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,768,0.021976000732845728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,512,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,256,0.003142222141226133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,512,0.01105955574247572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,512,0.020941333638297185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,256,0.019645333290100098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,128,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,64,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,4096,128,0.01854933301607768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,4096,32,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,64,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,256,0.011612444288200803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,4096,32,0.00998933364947637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,65536,0.07619111405478583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,51200,0.05858133236567179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,65536,0.06623022423850165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,65536,0.12066845099131267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,51200,0.09972533252504136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,16384,0.024591111474566992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,51200,0.05404622356096903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,16384,0.024118221468395654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,12288,0.018535999788178336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,16384,0.05016977919472588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,12288,0.02108088963561588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,10240,0.016142222616407607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,10240,0.023078221413824294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,12288,0.04355644517474704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,10240,0.039945777919557356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,8192,0.014367999302016364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,8192,0.021430222524536982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,7168,0.013753777576817406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,8192,0.036890665690104164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,7168,0.019723556107944913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,6144,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,7168,0.03564711080657111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,6144,0.01868799991077847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,5120,0.010913777682516309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,5120,0.01752711170249515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,6144,0.034084445900387235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,4096,0.009159999589125315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,5120,0.03173333406448364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,4096,0.028493334849675495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,3584,0.008849778109126622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,4096,0.016673778494199116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,3584,0.014764444695578681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,3072,0.008225777910815345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,3584,0.029688000679016113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,3072,0.013878222141000958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,2560,0.007256000406212277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,3072,0.02685866587691837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,2560,0.01351199961370892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,2048,0.005583111196756363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,2560,0.026240888569090102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,2048,0.012471111284361945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,1536,0.004524444540341695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,1536,0.011786667009194693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,2048,0.025261332591374714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,1024,0.003959111041492886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,1536,0.02309688925743103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,768,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,1024,0.021729777256647747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,1024,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,768,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,512,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,768,0.021379555265108745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,512,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,256,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,512,0.022322666313913133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,128,0.0029253332565228143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,256,0.010233778092596266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,256,0.020609777834680345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,64,0.0026640000028742682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3584,128,0.018258665998776753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3584,32,0.002875555513633622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,128,0.011319110790888468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,65536,0.06287733051511976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,32,0.010063110954231685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,65536,0.0632773306634691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,51200,0.05007911059591505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,65536,0.12065777513715957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,51200,0.051776889297697276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,16384,0.020993777447276648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,51200,0.09961511029137506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3584,64,0.011109333071443768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,16384,0.047862221797307335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,16384,0.024000888069470722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,12288,0.017106667160987854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,10240,0.016673778494199116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,12288,0.020560888780487906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,12288,0.043142222695880465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,10240,0.04007644454638163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,8192,0.014456889695591398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,10240,0.02351377738846673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,7168,0.012500444220172035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,8192,0.020865778128306072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,8192,0.036766221125920616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,6144,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,7168,0.019261333677503798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,7168,0.035029331843058266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,5120,0.009753777749008602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,6144,0.01831022236082289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,6144,0.03361777795685662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,5120,0.017090666625234816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,4096,0.00870400004916721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,5120,0.03253422180811564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,3584,0.00828888929552502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,4096,0.029508444997999404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,3584,0.027168888184759352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,3584,0.013703111145231458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,3072,0.007128888534175024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,3072,0.026118222210142348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,3072,0.013796444568369122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,2560,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,2560,0.013383111192120446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,2560,0.02512888941499922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,2048,0.005404444618357553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,4096,0.01533422205183241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,2048,0.012757333616415659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,2048,0.023392889234754775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,1536,0.004767111192146937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,1536,0.0118568887313207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,1024,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,1536,0.0244195560614268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,1024,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,768,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,1024,0.02292977770169576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,768,0.011125333607196808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,512,0.0032622222271230486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,768,0.02276266614596049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,512,0.011034666664070554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,256,0.0032426667296224167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,512,0.022322666313913133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,256,0.010071111222108206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,128,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,256,0.02159822152720557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,128,0.01076533314254549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,64,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,3072,128,0.021334222621387903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,3072,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,64,0.010444444086816577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,65536,0.055784887737698026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,3072,32,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,65536,0.06063110960854424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,51200,0.04572711057133145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,65536,0.12100799878438313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,51200,0.04971111151907179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,16384,0.019632889164818656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,16384,0.023062222533755835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,51200,0.10002488560146755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,12288,0.015094222293959724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,16384,0.047798222965664335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,12288,0.04134755664401584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,12288,0.020599999361568026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,10240,0.014142221874660917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,10240,0.03981066743532816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,10240,0.02148444453875224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,8192,0.01891377733813392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,8192,0.03685333331425985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,7168,0.011085333095656501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,7168,0.018292443619834054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,8192,0.011901333100265928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,6144,0.010760889285140567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,7168,0.03532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,6144,0.01715022159947289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,6144,0.033977776765823364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,5120,0.009830222361617619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,5120,0.029812445243199665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,5120,0.015782222151756287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,4096,0.008351999852392409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,4096,0.027505778604083594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,3584,0.007656888829337225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,4096,0.014172444740931192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,3584,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,3072,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,3584,0.02817866537306044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,3072,0.013157332936922709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,3072,0.026857778429985046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,2560,0.01311288856797748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,2560,0.02647999922434489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,2048,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,2048,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,2560,0.006248000181383557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,1536,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,1536,0.01146311147345437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,1024,0.0038986665507157645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,1024,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,1536,0.023433778021070693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,768,0.003597333199448056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,1024,0.021060445242457922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,768,0.011438222395049201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,768,0.020999999509917364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,512,0.0032142222755485107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,512,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,2048,0.025067556235525344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,512,0.02137600051032172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,256,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,128,0.0028151110228565005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,256,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,256,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,128,0.0107351111041175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,64,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2560,32,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2560,128,0.019640889432695176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,64,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,65536,0.045816000964906484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2560,32,0.011196444431940714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,65536,0.05842577748828464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,51200,0.03811733259095086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,65536,0.12052799595726861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,51200,0.04812533325619168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,16384,0.015299555328157214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,51200,0.09983733627531265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,16384,0.02271555529700385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,12288,0.013016000390052795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,16384,0.04808977908558316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,12288,0.019973332683245342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,10240,0.013239999612172445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,10240,0.019699555304315355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,10240,0.039628442790773176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,8192,0.011891555454995898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,8192,0.018280888597170513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,8192,0.03607111175855001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,7168,0.011304000185595619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,12288,0.04282755653063456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,7168,0.017039111918873258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,6144,0.010465777582592435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,7168,0.035106665558285184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,6144,0.015497777197096082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,6144,0.0319199992550744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,5120,0.009720889230569204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,5120,0.029612445169025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,5120,0.014477334088749357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,4096,0.008044444852405125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,4096,0.013804444836245643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,4096,0.027525333894623652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,3584,0.007746666669845581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,3584,0.013192888763215808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,3072,0.006711111300521427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,3584,0.02808177802297804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,3072,0.012814222110642327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,2560,0.005611555443869696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,3072,0.026729777455329895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,2560,0.012496000362767113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,2560,0.02632177703910404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,2048,0.012723555167516073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,2048,0.025379555092917547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,1536,0.004564444637960858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,1536,0.011892444557613797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,2048,0.004939555707905027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,1024,0.0038808890514903595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,1536,0.02450399928622776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,1024,0.011499555574523078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,1024,0.022729777627521094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,768,0.00360977773865064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,768,0.0210524449745814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,768,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,512,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,512,0.019813333948453266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,512,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,256,0.010384000009960597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,256,0.021751999855041504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,128,0.0028195555011431375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,2048,128,0.018619555566045973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,64,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,128,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,2048,32,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,64,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,65536,0.03618844350179037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,2048,32,0.011121778024567498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,65536,0.05654133028454251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,51200,0.029159112109078303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,65536,0.12082044283548991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,51200,0.04621244470278422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,16384,0.014737778239780001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,51200,0.1001688904232449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,16384,0.02218666672706604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,12288,0.012093333734406365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,16384,0.04900000161594815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,12288,0.019683554768562317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,10240,0.011348444554540845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,12288,0.0427706672085656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,10240,0.018287110659811232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,10240,0.039536890056398176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,8192,0.016604445046848722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,8192,0.03651644455062018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,7168,0.01034755590889189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,7168,0.0158897770775689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,8192,0.010127999716334874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,7168,0.03305066625277201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,6144,0.009568000005351173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,6144,0.015120888749758402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,5120,0.00872444444232517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,5120,0.014504889647165934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,5120,0.030060443613264296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,4096,0.007629333270920648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,4096,0.013864888913101621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,6144,0.031303998496797346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,3584,0.00721688899728987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,4096,0.028216888507207234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,3584,0.013086222112178802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,3072,0.006333333336644703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,3584,0.028190222051408555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,3072,0.012695999609099494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,3072,0.026682666606373254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,2560,0.005631111148330901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,2560,0.024456888437271118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,2560,0.012826667063766055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,2048,0.004966222163703707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,2048,0.02309155629740821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,2048,0.012540444731712341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,1536,0.004535111288229625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,1536,0.022701332966486614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,1024,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,1536,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,1024,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,1024,0.020587555236286588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,768,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,768,0.011134222149848938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,768,0.020695999264717102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,512,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,512,0.010945777926180096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,512,0.019880889190567862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,256,0.0029582221888833572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,256,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,256,0.019620445039537217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,128,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,128,0.010689777632554373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1536,128,0.01827555563714769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,64,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,64,0.010680889089902243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1536,32,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1536,32,0.010086221827401055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,65536,0.025242666403452556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,65536,0.05442577931616041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,65536,0.11991643905639648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,51200,0.021401777863502502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,51200,0.04438755578464932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,51200,0.09876710838741726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,16384,0.011490667031870948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,16384,0.021712000171343487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,16384,0.04661155409283108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,12288,0.010807111031479306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,12288,0.018499554859267343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,12288,0.04004266526963975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,10240,0.010016889207892949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,10240,0.017272000511487324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,10240,0.03829066620932685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,8192,0.008986666798591614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,8192,0.015800888339678448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,8192,0.03394577900568644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,7168,0.009480888644854227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,7168,0.015138667490747241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,6144,0.008605333666006723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,7168,0.03341155582004123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,6144,0.014548444085650973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,5120,0.007636444436179266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,6144,0.03090755475891961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,5120,0.014159111513031853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,5120,0.02988977895842658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,4096,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,3584,0.006578666468461354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,4096,0.02745955520206028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,3584,0.013129777378506131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,3072,0.006237333433495627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,3584,0.026729777455329895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,3072,0.012833777401182385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,2560,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,3072,0.025768889321221247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,2560,0.02511288887924618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,4096,0.0075502221783002215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,2048,0.006342222293217977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,2048,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,2048,0.023676445086797077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,1536,0.005599110904667113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,1536,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,1536,0.022322666313913133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,1024,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,1024,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,1024,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,768,0.0042133331298828125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,768,0.02071288890308804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,768,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,512,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,2560,0.012490666574902005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,256,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,512,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,512,0.019655111763212416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,128,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,256,0.011026666396194033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,256,0.02032355633046892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,64,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,128,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,1024,32,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,1024,128,0.01927199959754944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,65536,0.023735110958417256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,64,0.011031111081441244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,1024,32,0.009714666340086196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,51200,0.019923556182119582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,65536,0.05283288823233711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,65536,0.11945333745744492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,51200,0.04329599936803182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,16384,0.010617777705192566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,51200,0.09888088703155518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,12288,0.010650667051474253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,16384,0.020303999384244282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,16384,0.04727111260096232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,12288,0.018152889278199937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,12288,0.03936000002755059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,10240,0.01684533390733931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,8192,0.008293333152929941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,10240,0.0370124446021186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,8192,0.015815999772813585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,7168,0.008610666626029545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,7168,0.01514488955338796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,10240,0.009337777892748514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,7168,0.0319199992550744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,6144,0.00797333319981893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,8192,0.03361599975162082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,6144,0.014530667000346713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,5120,0.007327111230956183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,6144,0.030198223061031763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,5120,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,5120,0.028888887829250757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,4096,0.006983111302057902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,4096,0.013172444370057849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,3584,0.006928000185224745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,4096,0.027446221974160936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,3584,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,3584,0.02679288884003957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,3072,0.006255110932721033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,3072,0.025387555360794067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,3072,0.012451555165979596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,2560,0.005325333525737126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,2560,0.012148444851239523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,2560,0.02431022293037838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,2048,0.00490133340160052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,2048,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,1536,0.004948444250557157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,2048,0.023687111006842718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,1536,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,1536,0.022999111149046157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,1024,0.004368888835112254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,1024,0.02032977839310964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,1024,0.011211555865075854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,768,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,768,0.02092711130777995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,768,0.01072177787621816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,512,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,512,0.019828443725903828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,512,0.010771555205186209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,256,0.0032400000426504347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,256,0.0196079992585712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,256,0.01038044442733129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,128,0.010409778190983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,64,0.002536000063021978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,64,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,32,0.0025555555605226094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,768,128,0.021538666552967493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,65536,0.018710222509172227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,768,32,0.010424888796276517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,768,128,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,65536,0.05295377638604906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,65536,0.11913600232866074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,51200,0.04336355461014641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,51200,0.09816444582409328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,16384,0.010286221901575724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,16384,0.020228443874253165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,12288,0.00998222248421775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,16384,0.047741333643595375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,12288,0.017606221967273288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,10240,0.008749333520730337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,12288,0.04082399937841628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,51200,0.017460443907313876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,10240,0.01663644446267022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,8192,0.007735110819339752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,10240,0.03745511174201965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,7168,0.008796444369686974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,8192,0.03366311060057746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,8192,0.015554666519165039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,6144,0.008387555678685507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,7168,0.03219555483924018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,7168,0.015485333071814643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,5120,0.007685333490371704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,6144,0.014515555567211576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,6144,0.030255999830034044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,4096,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,5120,0.029250668154822454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,5120,0.01425244410832723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,3584,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,4096,0.013173333472675748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,4096,0.027464000715149775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,3072,0.006326222171386083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,3584,0.026175111532211304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,3584,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,2560,0.006284444282452266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,3072,0.025458667013380263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,2560,0.02407111061943902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,2048,0.005952888892756567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,2048,0.011691555380821228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,2048,0.023031999667485554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,1536,0.004263111286693149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,1536,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,3072,0.012468444804350534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,1536,0.022283555732833013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,1024,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,2560,0.01273333364062839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,1024,0.021673777037196692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,768,0.003744889050722122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,768,0.010738666686746808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,512,0.003354666754603386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,768,0.022837332553333704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,512,0.010959111154079437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,256,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,512,0.022433777650197346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,1024,0.010747555229398938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,256,0.019738665885395475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,128,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,128,0.010403555300500657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,512,128,0.018331555856598746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,64,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,256,0.01038755559258991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,512,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,32,0.010406222608354356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,65536,0.0166311115026474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,65536,0.051078223519855075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,65536,0.11901777320437962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,51200,0.015508444772826301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,512,64,0.010176888770527309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,51200,0.04046488801638285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,51200,0.0972835554016961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,16384,0.009111111362775167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,16384,0.020022221737437777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,16384,0.04619110955132378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,12288,0.007688889073001013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,12288,0.038831111457612776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,12288,0.017595556047227647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,10240,0.007057777709431118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,10240,0.01683822274208069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,10240,0.03694933321740892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,8192,0.0063288890653186375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,8192,0.015476443701320224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,8192,0.03362755642996894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,7168,0.01517955627706316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,7168,0.03228088882234361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,6144,0.006685333533419504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,6144,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,6144,0.03126666612095303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,5120,0.005987555616431766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,7168,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,5120,0.029137776957617864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,5120,0.01348533315791024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,4096,0.00564444437623024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,4096,0.027469333675172593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,4096,0.013103110922707452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,3584,0.005629333357016246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,3584,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,3584,0.02602133320437537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,3072,0.005422222117582957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,3072,0.012520888613329994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,2560,0.006011555592219035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,3072,0.025762667258580525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,2560,0.012166221936543783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,2560,0.025106666816605464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,2048,0.005566222386227713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,2048,0.023046221998002794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,2048,0.012094222009181976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,1536,0.004274666723277834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,1536,0.022336888644430373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,1536,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,1024,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,1024,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,768,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,768,0.010814222196737925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,768,0.021328000558747187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,512,0.0032417778339650896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,512,0.010992888775136737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,512,0.020960888928837244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,1024,0.020644444558355544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,256,0.003030222323205736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,128,0.002936888900068071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,256,0.019640889432695176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,256,0.010455999937322406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,64,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,128,0.011240888800885944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,256,128,0.018479110466109384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,256,32,0.003026666740576426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,65536,0.014329777823554145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,32,0.009408888717492422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,256,64,0.010793777803579966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,51200,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,65536,0.04818400078349643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,65536,0.11780533525678848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,16384,0.007715555528799693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,51200,0.03943822118971083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,51200,0.09657510783937241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,12288,0.007756444315115611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,16384,0.01996977792845832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,16384,0.04606399933497111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,12288,0.039092444711261325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,12288,0.01832444469134013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,10240,0.00701422244310379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,8192,0.0063546668324205614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,10240,0.016599999533759225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,10240,0.0380364457766215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,7168,0.006614222294754452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,8192,0.033639109796947904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,8192,0.015533333023389181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,6144,0.005972444597217772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,7168,0.015277332729763456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,7168,0.031934221585591636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,5120,0.005673777725961473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,6144,0.031192888816197712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,6144,0.014501333236694336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,4096,0.005306666509972678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,5120,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,5120,0.029863112502627905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,4096,0.013213333156373767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,4096,0.02811644474665324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,3584,0.005601777798599667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,3584,0.026104888982243005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,3584,0.013075555364290873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,3072,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,3072,0.01275644451379776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,3072,0.025644444757037695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,2560,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,2560,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,2048,0.005305777821275923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,2048,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,2048,0.023103111320071753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,1536,0.004591111093759537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,1536,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,1536,0.02253066665596432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,1024,0.003583111282851961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,2560,0.025395555628670588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,1024,0.010718222293588849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,768,0.003378666730390655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,1024,0.02202400068442027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,768,0.010795555180973478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,768,0.021959111094474792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,512,0.0031591111587153543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,512,0.01965777741538154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,512,0.011227555572986603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,256,0.0028168888141711554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,256,0.010769777827792697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,256,0.019702222612169053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,128,0.010412444670995077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,12,128,128,0.018266666266653273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,64,0.0025964443468385274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,128,32,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,64,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,65536,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,128,32,0.009447111023796929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,51200,0.010319111247857412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,65536,0.04797244403097364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,16384,0.0069297779765393995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,51200,0.03906133439805772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,12288,0.00658044425977601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,16384,0.020404444800482858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,12288,0.017779555585649278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,10240,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,8192,0.006607999818192587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,10240,0.016533333394262526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,7168,0.005588444570700328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,8192,0.015816888875431485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,6144,0.006270222365856171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,7168,0.014879110786649915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,5120,0.0052622221410274506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,6144,0.01480977733929952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,4096,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,5120,0.013855111267831592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,3584,0.005612444546487596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,4096,0.013456000222100152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,3072,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,3584,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,2560,0.005560889012283749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,3072,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,2048,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,2560,0.012104888757069906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,1536,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,2048,0.012065778175989786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,1024,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,1536,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,768,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,1024,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,512,0.0031840000301599503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,768,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,256,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,512,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,128,0.0028231110837724474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,256,0.009931555224789513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,64,0.0024986666523747975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,128,0.011344888971911537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,64,32,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,64,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,65536,0.010716444088353051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,64,32,0.01036444471942054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,51200,0.010024888647927178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,65536,0.0472942226462894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,16384,0.006487111250559489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,51200,0.03947644432385763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,12288,0.007325333025720384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,16384,0.019920888874265883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,10240,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,12288,0.01819022165404426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,8192,0.006962666908899943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,10240,0.016549333930015564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,7168,0.005616888817813661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,8192,0.015806222955385845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,6144,0.0063164445261160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,7168,0.014872888724009195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,5120,0.005220444252093633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,6144,0.014500444134076437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,4096,0.005569777968857024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,5120,0.013790222505728403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,4096,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,3584,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,3584,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,3072,0.005277333160241445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,3072,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,2560,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,2560,0.012151111331250934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,2048,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,2048,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,1536,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,1536,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,1024,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,1024,0.010692444112565784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,768,0.003527111063400904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,768,0.011162666810883416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,512,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,512,0.010391111175219217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,256,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,256,0.00981777740849389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,128,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,128,0.010405333505736457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,64,0.00253866674999396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,64,0.010097777677906884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,12,32,32,0.002464888824356927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,12,32,32,0.009352888498041365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,51200,0.9514239629109701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,51200,0.4948435359530979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,16384,0.167684449089898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,16384,0.33659911155700684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,12288,0.22589333852132162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,12288,0.12971733676062688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,10240,0.18923466735415986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,10240,0.1254648897382948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,8192,0.15219377146826849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,7168,0.13392089472876653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,8192,0.13101422786712646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,6144,0.11591733826531304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,7168,0.11858044730292426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,6144,0.08750577767690022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,5120,0.09835022025638157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,4096,0.07792177465226915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,5120,0.08949333429336548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,3584,0.06904621918996175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,4096,0.06674310896131727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,3584,0.058724443117777504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,3072,0.059839997026655406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,2560,0.05062399970160591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,3072,0.051875554853015475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,2048,0.04142311215400696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,2560,0.05043733451101515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,1536,0.03218577636612786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,2048,0.041843556695514254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,1024,0.02314844396379259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,1536,0.03154133425818549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,768,0.018556444181336295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,1024,0.025435555312368605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,512,0.013277333643701343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,512,0.018216000662909616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,768,0.022566222482257422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,256,0.008839999636014303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,128,0.006361777583758037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,256,0.01632355484697554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,128,0.01352533366945055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,64,0.01350044459104538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,32,0.005616888817813661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,65536,32,0.014871110518773397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,65536,64,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,65536,1.0273218154907227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,65536,0.5228248702155219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,51200,0.7982915772332085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,51200,0.4007999897003174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,16384,0.2748897870381673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,16384,0.13702310456169978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,12288,0.1995013289981418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,12288,0.10674133565690781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,8192,0.13487466176350912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,10240,0.16628266705407038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,7168,0.12023644977145725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,8192,0.09242755836910671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,10240,0.10231822066836888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,7168,0.08289421929253472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,5120,0.08502133025063409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,6144,0.07283466392093234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,4096,0.07145600186453925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,5120,0.0747928884294298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,3584,0.06214844518237644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,4096,0.055240889390309654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,3584,0.051009777519438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,3072,0.05339822173118591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,2560,0.04551644457711113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,3072,0.044249776336881846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,2048,0.03803733322355483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,2560,0.04072711202833388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,1536,0.029608001311620075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,6144,0.10238399770524766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,2048,0.03667466508017646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,1024,0.020425778296258714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,768,0.01590399940808614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,1024,0.022384888596004907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,1536,0.028220444917678833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,512,0.011792889071835412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,768,0.020690666304694284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,512,0.017287999391555786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,256,0.007385777930418651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,128,0.003986666599909465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,256,0.01406044430202908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,64,0.003584889074166616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,128,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,51200,32,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,64,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,51200,32,0.015053333507643806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,65536,0.33073245154486763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,65536,0.1833324432373047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,51200,0.2594533231523302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,16384,0.098036441538069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,51200,0.14014577865600586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,12288,0.06989866495132446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,16384,0.054735110865698926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,12288,0.04483288857671949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,10240,0.060421334372626416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,8192,0.04303466611438327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,10240,0.05224177903599209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,7168,0.037738667594061956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,8192,0.04377955529424879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,6144,0.03310399916436937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,7168,0.04184977875815498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,5120,0.028618666860792372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,6144,0.03505333264668783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,4096,0.023248000277413264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,5120,0.03262933426433139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,3584,0.02103288968404134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,4096,0.02732622292306688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,3072,0.018407111366589863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,2560,0.015832000308566622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,3584,0.025428444147109985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,2048,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,3072,0.022788445154825848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,1536,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,2560,0.02179733415444692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,1024,0.008323555191357931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,2048,0.019912888606389362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,768,0.0069475554757648045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,1536,0.017156443662113614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,1024,0.015488889482286243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,512,0.004947555561860402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,256,0.003681777666012446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,768,0.014183110660976835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,128,0.0029075555503368378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,512,0.013214222258991666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,256,0.011811555259757571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,16384,32,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,128,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,64,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,16384,32,0.011751111182901593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,65536,0.24529067675272623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,65536,0.1550497743818495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,51200,0.1938800017038981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,16384,0.06868533293406169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,12288,0.05065422256787618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,51200,0.12131467130449082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,10240,0.04390666551060147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,16384,0.048399110635121666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,12288,0.03982399900754293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,8192,0.036489778094821505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,10240,0.04500800040033129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,7168,0.03250577714708116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,6144,0.028936889436509874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,7168,0.0370782216389974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,5120,0.023748444186316595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,6144,0.029564443561765883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,4096,0.01984177695380317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,5120,0.028352889749738906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,3584,0.017542221479945712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,4096,0.024686222275098164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,3072,0.015558222929636637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,3584,0.02347822156217363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,2560,0.01348977784315745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,3072,0.02088622252146403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,8192,0.0377439988984002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,2048,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,2560,0.021020443903075323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,1536,0.00978577799267239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,1024,0.00776977754301495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,2048,0.01867644488811493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,768,0.006542222367392645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,1536,0.016169778174824186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,1024,0.013879111243618859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,768,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,512,0.004252444538805219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,256,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,128,0.0032391111469931076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,512,0.012668444050682915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,64,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,256,0.012516444755925072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,128,0.011789333489206104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,12288,32,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,32,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,65536,0.20579289065466988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,65536,0.1494613356060452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,51200,0.16032444106207952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,51200,0.11670311292012532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,16384,0.06375822093751696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,12288,64,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,12288,0.04554133282767402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,16384,0.04711822337574429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,10240,0.038455112112893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,12288,0.03861155443721347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,8192,0.03180355495876736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,8192,0.032955557107925415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,7168,0.028864890336990356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,7168,0.02953333324856228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,6144,0.025968889395395916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,10240,0.03821333249409994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,6144,0.027256000373098586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,5120,0.022601778308550518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,4096,0.017882666654056974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,5120,0.02424888809521993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,3584,0.01594311164485084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,4096,0.021360889077186584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,3072,0.014292443792025248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,3584,0.020766221814685397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,3072,0.01959377858373854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,2560,0.012672888735930124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,2560,0.01834844383928511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,2048,0.011033777561452655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,1536,0.009374221993817223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,2048,0.01706044375896454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,1024,0.00738933351304796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,1536,0.014120888378885059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,768,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,1024,0.013474666410022311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,512,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,768,0.012366222010718452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,512,0.012500444220172035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,256,0.0037617778612507712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,128,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,256,0.012085333466529846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,128,0.010846222440401712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,64,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,10240,32,0.003162666741344664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,10240,32,0.01074399964676963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,65536,0.16349155373043484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,51200,0.1289057731628418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,65536,0.09961155388090347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,16384,0.05073066552480062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,16384,0.03493422269821167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,51200,0.07851822508705987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,12288,0.03268533282809787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,10240,0.02826311190923055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,10240,0.031181332137849595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,8192,0.023216888308525085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,8192,0.026837334036827087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,7168,0.02093244426780277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,7168,0.02509333358870612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,6144,0.018615111708641052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,6144,0.022834666901164587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,5120,0.015936889582210116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,12288,0.029690666331185236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,5120,0.02095111045572493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,4096,0.013460444079505073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,3584,0.012226666841242047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,4096,0.01889244384235806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,3072,0.010971555279360877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,3584,0.018315555320845712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,3072,0.017317333155208163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,2560,0.009842666486899057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,2048,0.008461332983440822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,2560,0.01663822266790602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,1536,0.007024889190991719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,1536,0.013238222234778933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,1024,0.0052586665583981415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,2048,0.014928888943460254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,768,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,1024,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,512,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,512,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,768,0.012219555675983429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,256,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,128,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,128,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,64,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,256,0.011124444504578909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,8192,32,0.0029440000653266907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,64,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,65536,0.1540071169535319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,8192,32,0.011136888629860349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,51200,0.11416444513532852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,65536,0.09105510844124688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,16384,0.042391998900307544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,12288,0.03323644399642944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,16384,0.03299999899334378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,51200,0.07188889053132799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,10240,0.02887644370396932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,12288,0.027415111660957336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,8192,0.02402044501569536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,7168,0.021941334009170532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,8192,0.025711999999152288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,10240,0.028845333390765723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,7168,0.024363555841975745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,6144,0.022111111217074927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,5120,0.015027556154463025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,5120,0.020351111888885498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,4096,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,4096,0.018585777944988675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,3584,0.011210666762457954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,6144,0.016912889149453905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,3584,0.018046221799320646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,3072,0.010016000105275048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,2560,0.009095999929640029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,3072,0.01687022215790219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,2048,0.008045333127180735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,2560,0.0163128889269299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,2048,0.014763555592960782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,1536,0.006895111252864202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,1024,0.005024000174469418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,1536,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,768,0.003548444559176763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,768,0.012279111478063794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,512,0.003253333270549774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,1024,0.012350222302807702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,256,0.003273777870668305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,512,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,128,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,128,0.011492444409264458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,256,0.011121778024567498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,64,0.0026177778426143858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,7168,32,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,32,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,7168,64,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,65536,0.12830844190385607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,51200,0.09522222148047553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,65536,0.08243466748131646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,16384,0.03684444559945001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,51200,0.06590133243136935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,12288,0.02805333336194356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,12288,0.02567555507024129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,16384,0.030844443374209937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,10240,0.024772443705134924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,8192,0.020996444755130347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,10240,0.02680533296532101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,8192,0.024591111474566992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,7168,0.01889955500761668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,6144,0.01699199941423204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,7168,0.023790222075250413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,5120,0.015320888823933072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,6144,0.021338666478792827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,5120,0.01992266707950168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,4096,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,3584,0.01075111081202825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,4096,0.01795111099878947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,3072,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,3584,0.01790488925245073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,3072,0.016065778003798593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,2560,0.00871288859181934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,2048,0.007700444095664554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,2560,0.015582222077581616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,1536,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,2048,0.014222222897741528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,1024,0.0041617775956789655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,1536,0.01291911138428582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,768,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,768,0.012161778079138862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,512,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,1024,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,256,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,512,0.012076444096035428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,256,0.010847999817795224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,128,0.011295110815101199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,64,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,6144,32,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,64,0.011197333534558615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,6144,32,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,65536,0.10667555862002903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,51200,0.08768177694744533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,65536,0.07499111360973783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,16384,0.03456622362136841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,51200,0.06036533249749077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,12288,0.025768889321221247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,10240,0.022035555707083807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,16384,0.02854577700297038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,12288,0.02342133389578925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,8192,0.01924000018172794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,7168,0.01722666621208191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,10240,0.0244186669588089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,8192,0.022854222191704646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,6144,0.015428443749745687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,5120,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,7168,0.02269777821169959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,6144,0.01975911193423801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,4096,0.010470222267839642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,3584,0.00981422182586458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,5120,0.018812444474962022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,4096,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,3072,0.008776889079146916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,2560,0.007983999947706858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,3584,0.017536888519922893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,3072,0.01482488877243466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,2048,0.0069573331210348345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,2560,0.014896889527638754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,1536,0.005260444349712796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,2048,0.013820444544156393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,1024,0.003979555434650845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,1536,0.012429333395428128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,768,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,1024,0.012421333127551608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,512,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,768,0.011863999896579318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,512,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,256,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,128,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,256,0.011506666739781698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,64,0.0025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,128,0.010788444843557147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,5120,32,0.002557333351837264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,64,0.011113777756690979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,65536,0.08627111381954616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,5120,32,0.010746666954623329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,51200,0.06699289215935601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,65536,0.06803644365734525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,16384,0.028026666906144884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,16384,0.025451555848121643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,51200,0.055480887492497764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,12288,0.02065955599149068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,10240,0.018535999788178336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,12288,0.021692444880803425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,10240,0.022988445229000513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,8192,0.021196444829305012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,7168,0.013844444519943662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,7168,0.021011556188265484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,6144,0.012513777448071374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,6144,0.018592889110247295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,5120,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,5120,0.017539555827776592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,8192,0.015440889530711703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,4096,0.009402666654851701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,3584,0.008919999831252629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,4096,0.016552888684802584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,3584,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,3072,0.007985778152942657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,2560,0.0075662218862109715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,3072,0.013434666726324292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,2048,0.005952000204059813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,2560,0.014529777897728814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,2048,0.013163555827405719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,1536,0.004661333229806689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,1024,0.003999999827808804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,1536,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,768,0.003600888782077365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,768,0.01204533295498954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,512,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,1024,0.011953777737087674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,256,0.003035555490189128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,512,0.011725333001878528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,128,0.0028799999919202593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,256,0.01072355525361167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,64,0.002508444504605399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,128,0.01131199962562985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,4096,32,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,32,0.010407110883129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,4096,64,0.011205332974592844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,65536,0.07594755623075697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,51200,0.058640890651279025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,65536,0.06571822033988105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,16384,0.024028443627887305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,51200,0.053364445765813194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,12288,0.01847999956872728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,16384,0.024355555574099224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,12288,0.021183111601405676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,10240,0.0159608887301551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,8192,0.014137778017255994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,10240,0.022105778257052105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,7168,0.013558222187889947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,8192,0.021664000219768945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,7168,0.019983111156357657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,6144,0.017954667409261067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,5120,0.010836444795131683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,5120,0.01705422169632382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,4096,0.009101333717505137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,4096,0.015275556180212231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,3584,0.008431110945012834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,3584,0.014582221706708273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,6144,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,3072,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,2560,0.00701422244310379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,2560,0.014106666048367819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,2048,0.005631999837027655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,2048,0.013217777841620974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,1536,0.0046977777447965415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,3072,0.007830222447713217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,1024,0.0038648889296584656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,768,0.0036764442920684814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,1536,0.01221244451072481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,1024,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,512,0.00332266671790017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,768,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,256,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,512,0.011767110890812345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,128,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,256,0.010998222563001843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,128,0.011133333047231039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3584,32,0.0028124445428450904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,64,0.010676444404655032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,65536,0.06278666522767809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3584,32,0.010401777923107147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,51200,0.05044355657365587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,65536,0.06264977984958224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,16384,0.021216000119845074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,51200,0.0513555539978875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,12288,0.016576000385814242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,16384,0.023511999183230933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,10240,0.01717155509524875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,12288,0.020636444290479023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,8192,0.014542222023010254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,10240,0.021680000755521987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,7168,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,8192,0.0192239996459749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,6144,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,7168,0.01869422197341919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,5120,0.01015377789735794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,6144,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,4096,0.008664888640244802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,5120,0.016539555456903245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,3584,0.008348444269763099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,4096,0.014085332552591959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,3072,0.007240888973077138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,3584,0.014533332652515836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,2560,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,3072,0.013095999757448832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,2048,0.004878222114510006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,2560,0.01331911070479287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,1536,0.004553777890072929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,2048,0.012450666891203986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,1024,0.0038613333470291565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,1536,0.012154666913880242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,768,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,1024,0.012055111428101858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,512,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,768,0.011353777514563667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,512,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,128,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,256,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,64,0.0027875554644399216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,128,0.01146044416560067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,3072,32,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,64,0.010431110858917236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,65536,0.05577244361241659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,3072,32,0.01036088913679123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,51200,0.04587911234961616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,65536,0.06048533651563856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,16384,0.019716444942686293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,51200,0.04935466580920749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,12288,0.015074667003419666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,16384,0.023009777069091797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,10240,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,12288,0.020290666156344943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,8192,0.01198755535814497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,10240,0.020433778564135235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,7168,0.011184888581434885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,8192,0.018945778409639995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,6144,0.010479110810491772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,7168,0.017771555317772757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,5120,0.009661333428488838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,6144,0.01594666639963786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,4096,0.008325333396593729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,5120,0.014868444866604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,3584,0.007307555940416124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,4096,0.013752000199423896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,3072,0.006636444479227066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,3584,0.013795555465751223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,2560,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,3072,0.013085333009560903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,2048,0.00490044429898262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,2560,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,1536,0.00452622233165635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,2048,0.012392000191741519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,1024,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,768,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,1536,0.012031111452314587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,512,0.003491555651028951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,768,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,256,0.0031386667655573953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,512,0.011348444554540845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,128,0.0027919999427265594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,256,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,64,0.002491555487116178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,128,0.010391111175219217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2560,32,0.002540444541308615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,1024,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,64,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,65536,0.04599110947714912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2560,32,0.00965244405799442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,51200,0.037866665257347956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,65536,0.05814222494761149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,16384,0.015425778097576566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,51200,0.04763288961516487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,12288,0.012625777886973487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,16384,0.022682666778564453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,10240,0.013415111435784234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,12288,0.020104888412687514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,8192,0.011550222006109027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,10240,0.019165333774354722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,7168,0.010768888725174798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,7168,0.016099555624855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,6144,0.010370666782061258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,6144,0.0148062225845125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,5120,0.009732444253232744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,5120,0.014224888549910652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,4096,0.008307555483447181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,4096,0.013450666434235044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,8192,0.018600889378123812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,3584,0.007359111474619971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,3072,0.00600622221827507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,3584,0.01346577786737018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,2560,0.005551110953092575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,3072,0.01275111072593265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,2048,0.004912000149488449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,2560,0.012699555191728802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,1536,0.00423200014564726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,2048,0.012073777616024017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,1024,0.0038648889296584656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,1536,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,768,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,1024,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,512,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,768,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,256,0.003160888950030009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,512,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,128,0.0028311111446883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,256,0.011008888483047485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,64,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,128,0.011115555961926779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,2048,32,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,64,0.010458666417333813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,65536,0.03583200110329522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,2048,32,0.010334221853150262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,51200,0.02932888931698269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,65536,0.05645333396063911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,16384,0.015028445257080926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,51200,0.046408000919553966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,12288,0.01202666676706738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,16384,0.02206755512290531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,10240,0.01147377739350001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,12288,0.019612444771660697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,8192,0.010016000105275048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,10240,0.018188445104493033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,7168,0.010414222048388587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,8192,0.015835555063353646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,6144,0.00944800012641483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,7168,0.01517777807182736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,5120,0.008272888759771982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,6144,0.014808000789748298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,4096,0.007631111476156447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,5120,0.013973333769374423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,3584,0.006625777731339137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,4096,0.01348533315791024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,3072,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,3584,0.013471999930010902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,2560,0.005659555395444234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,3072,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,2048,0.005208888815508948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,2560,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,1536,0.00425866660144594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,2048,0.012376888758606382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,1024,0.0039048890272776284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,1536,0.012068444656001197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,768,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,1024,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,512,0.0033608888172441055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,768,0.011422221859296164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,256,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,512,0.011372444530328115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,128,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,256,0.011032000184059143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,64,0.002513777878549364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,128,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1536,32,0.002510222295920054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,64,0.010384000009960597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,65536,0.02514933380815718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1536,32,0.010039110978444418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,51200,0.02152266601721446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,65536,0.05395822061432732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,16384,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,51200,0.04384711053636339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,12288,0.010988444089889526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,16384,0.021670222282409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,10240,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,12288,0.017847999930381775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,8192,0.008999110923873054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,10240,0.01717066599263085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,7168,0.009283555878533257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,8192,0.01555288831392924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,6144,0.008376888930797577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,7168,0.015191111299726697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,5120,0.007807999849319458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,6144,0.014510222607188754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,5120,0.014173333843549093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,4096,0.0069360000391801195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,4096,0.013455111119482251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,3584,0.006587555425034628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,3584,0.013330666555298699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,3072,0.005897777775923411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,3072,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,2560,0.005544888890451855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,2560,0.012792000340090858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,2048,0.006304889089531369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,2048,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,1536,0.005552000055710475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,1536,0.01218488895230823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,1024,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,1024,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,768,0.003914666672547658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,768,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,512,0.0034968890249729156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,512,0.011704888608720569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,256,0.003173333282272021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,256,0.011501332951916588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,128,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,128,0.011015999648306104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,64,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,64,0.011029333704047732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,1024,32,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,65536,0.023583999938435022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,65536,0.05266399847136604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,51200,0.01907911068863339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,51200,0.043545779254701406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,16384,0.010462221999963125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,1024,32,0.010379555324713389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,16384,0.020743111769358318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,12288,0.010077333284748925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,10240,0.009355555805895064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,12288,0.01787644459141625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,10240,0.01687111126052009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,8192,0.008271111382378472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,7168,0.00865955568022198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,8192,0.015585778488053216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,6144,0.008040000167157915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,7168,0.01514933341079288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,5120,0.007705777883529663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,6144,0.014568888478808932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,4096,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,5120,0.014092443717850579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,3584,0.006830222076839871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,4096,0.01346666696998808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,3072,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,3584,0.01311555587583118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,3072,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,2560,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,2048,0.0048942222363419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,2048,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,1536,0.004885333279768626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,1536,0.012071111136012606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,1024,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,1024,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,768,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,2560,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,512,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,768,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,256,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,512,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,128,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,256,0.011439999772442712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,64,0.0025519999778932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,128,0.010363555616802638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,768,32,0.0025439999169773525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,64,0.011047999891969891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,65536,0.018826666805479262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,768,32,0.01033688916100396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,51200,0.016520889268981088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,65536,0.05211822191874186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,16384,0.009894222021102905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,51200,0.043172445562150746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,12288,0.009699555734793345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,16384,0.020297777321603563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,10240,0.008641777767075432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,12288,0.01787911189926995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,8192,0.007972444097201029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,10240,0.01684533390733931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,7168,0.0086986662613021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,8192,0.015475556254386902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,7168,0.01513066722287072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,6144,0.008064889245563084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,5120,0.007313778003056844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,6144,0.014517333772447375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,4096,0.0069288888739215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,5120,0.014156444205178155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,3584,0.006704000135262807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,4096,0.013410666750537025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,3072,0.0063742221229606206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,3584,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,2560,0.006080889039569431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,3072,0.012808889150619507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,2048,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,2560,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,1536,0.004295999805132548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,2048,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,1536,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,1024,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,768,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,1024,0.011704888608720569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,512,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,512,0.01145600030819575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,768,0.011674666570292579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,256,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,128,0.0028719999310043124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,128,0.011198221809334226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,256,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,64,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,512,32,0.002551111082235972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,32,0.010463111102581024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,65536,0.016353777713245816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,65536,0.04966844452752007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,51200,0.015511999527613321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,51200,0.040023111634784274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,16384,0.00868088917599784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,16384,0.02032444377740224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,12288,0.007738666401969061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,512,64,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,12288,0.017916444275114272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,10240,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,8192,0.006803555621041193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,10240,0.0169493340783649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,7168,0.006396444307433234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,8192,0.015983111328548856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,7168,0.01517599986659156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,6144,0.006983111302057902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,5120,0.007061333292060428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,6144,0.014544889330863953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,4096,0.00629688882165485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,5120,0.014312000738249885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,4096,0.013226666384273104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,3584,0.006133333262470033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,3072,0.005967111223273807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,3584,0.013154666456911298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,2560,0.006007110906971826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,2560,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,3072,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,2048,0.005945777727497949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,1536,0.004230222354332606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,1536,0.012105777859687805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,2048,0.012585777375433179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,1024,0.0038853333228164246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,768,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,768,0.011383111278216044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,1024,0.01165688865714603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,512,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,256,0.002926222152180142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,256,0.011505777637163797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,128,0.0026808888134029177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,128,0.011231111155615913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,64,0.002556444456179937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,64,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,256,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,32,0.010462221999963125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,65536,0.014146667387750415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,65536,0.04803466796875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,256,512,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,51200,0.012091555529170565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,16384,0.008031110796663497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,16384,0.020364445116784837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,51200,0.03985244366857741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,12288,0.007341333561473423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,10240,0.0069857777820693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,10240,0.016918222109476726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,12288,0.018235555953449674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,8192,0.006650666809744305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,7168,0.006311111152172089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,8192,0.015981333123313058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,7168,0.015190222197108798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,6144,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,5120,0.0060159998635451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,6144,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,4096,0.005730666634109285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,5120,0.014209777116775513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,4096,0.013453333742088742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,3584,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,3072,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,2560,0.005995555470387141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,3584,0.013248888982666863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,2048,0.005227555417352253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,2560,0.012800888882742988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,3072,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,1536,0.004286222159862518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,1024,0.0038559999730851916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,2048,0.011966221862369113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,768,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,1024,0.01149511088927587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,1536,0.012325333224402534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,512,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,256,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,768,0.010808889236715106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,128,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,256,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,64,0.0027644443843099806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,128,0.010101333260536194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,512,0.011877333124478659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,128,32,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,32,0.011040888726711273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,65536,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,65536,0.04771999849213494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,51200,0.010010666317409938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,51200,0.039677331844965615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,128,64,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,16384,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,12288,0.0074106670088238185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,16384,0.02033866610791948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,10240,0.00674044465025266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,12288,0.01834488908449809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,8192,0.006675555474228329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,10240,0.016898666818936665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,7168,0.006659555352396435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,8192,0.015885333220163982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,7168,0.015226667126019796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,6144,0.006314666734801398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,5120,0.006018666757477655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,6144,0.014820444915029736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,4096,0.00562755556570159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,5120,0.014185777968830533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,3584,0.006158222340875202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,4096,0.013479111095269522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,3072,0.005632888939645555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,3072,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,3584,0.013605333036846586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,2560,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,2048,0.005603555589914322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,2560,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,2048,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,1536,0.0046675557063685525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,1024,0.00388711111413108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,1536,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,768,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,1024,0.011810666984981961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,512,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,768,0.011761777930789523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,256,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,512,0.011762667033407422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,128,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,128,0.011223999990357293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,64,0.0029306666304667792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,64,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,64,32,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,32,0.011061333119869232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,64,256,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,65536,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,51200,0.010047111246320937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,65536,0.04766399992836846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,16384,0.006996444529957241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,51200,0.039678222603268094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,16384,0.020236444142129686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,12288,0.007349333001507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,10240,0.006803555621041193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,12288,0.017993777990341187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,8192,0.006728888799746831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,10240,0.01700711084736718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,7168,0.006671999891599019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,8192,0.01586666703224182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,6144,0.0064284445510970215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,7168,0.01548266741964552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,6144,0.01479288935661316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,5120,0.006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,4096,0.00564177789621883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,5120,0.014175110393100314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,3584,0.005659555395444234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,4096,0.013537777794731988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,3072,0.005696889013051987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,3584,0.013387555877367655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,2560,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,3072,0.013208889298968844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,2560,0.012810666528013019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,2048,0.005640888793600931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,1536,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,2048,0.012169777519173093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,1024,0.003927111211750242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,1536,0.012448000411192576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,768,0.0036186666952239144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,768,0.011534222298198275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,512,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,512,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,256,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,256,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,128,0.002934222213096089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,1024,0.01182755579551061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,64,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,128,0.011186666786670685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,11,32,32,0.0025395556456512874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,64,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,11,32,32,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,51200,0.4943902227613661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,51200,0.9493342505560981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,16384,0.33652978473239475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,16384,0.16734666294521758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,12288,0.24131467607286242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,12288,0.18276088767581514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,10240,0.18891110685136583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,8192,0.15197243955400255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,10240,0.12995999389224583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,8192,0.13215377595689562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,7168,0.13391289446089003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,7168,0.11343022187550862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,6144,0.11538310845692952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,6144,0.09515021906958686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,5120,0.0977591143714057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,5120,0.07809955543941922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,4096,0.07758933305740356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,4096,0.0676800012588501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,3584,0.06881778107749091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,3584,0.06329511271582709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,3072,0.05979910824033949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,3072,0.05924533473120796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,2560,0.05048888921737671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,2560,0.04559822214974297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,2048,0.04150044586923387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,2048,0.03720888826582167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,1536,0.03198666705025567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,1536,0.03307377629809909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,1024,0.02281422250800663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,1024,0.02576444380813175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,768,0.018156444032986958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,768,0.021967111362351313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,512,0.013256000147925483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,512,0.018632888793945312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,256,0.008681777450773451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,256,0.015807110402319167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,128,0.006254222244024277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,128,0.013839999834696451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,64,0.0055440002017551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,64,0.01479644411140018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,65536,32,0.00555644432703654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,65536,32,0.015464888678656684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,65536,1.0311128828260634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,65536,0.5221590995788574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,51200,0.7948791186014811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,16384,0.2759733464982775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,51200,0.39983733495076496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,12288,0.19918489456176758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,16384,0.13671555784013537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,12288,0.14681333965725368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,10240,0.1057279970910814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,10240,0.1657866636912028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,8192,0.13833510875701904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,7168,0.11740356021457249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,6144,0.10362578100628322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,8192,0.10526844527986313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,7168,0.09595021936628555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,5120,0.08519022332297431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,6144,0.07804177867041694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,4096,0.06953510973188612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,3584,0.06120889054404365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,4096,0.058111111323038735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,3072,0.05372088816430834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,3584,0.05344266692797343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,5120,0.06499199734793769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,2560,0.04435733291837904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,2048,0.03754666778776381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,3072,0.046241776810752026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,2560,0.039275556802749634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,1536,0.02983733349376255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,2048,0.03237688872549269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,1536,0.028675556182861328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,1024,0.02270488937695821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,768,0.016413332687483895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,512,0.011667555405033959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,768,0.01979111135005951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,256,0.0075351107451650835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,512,0.016897777716318767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,128,0.0038577777643998465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,256,0.01479466590616438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,64,0.0035386664999855887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,128,0.012710221939616732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,32,0.0035333335399627686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,64,0.014132445057233175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,51200,32,0.014506666196717156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,51200,1024,0.021151110529899597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,65536,0.3308417797088623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,51200,0.25891023212009007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,65536,0.18237155013614229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,16384,0.09675377607345581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,51200,0.1396204498079088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,16384,0.05475111140145195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,10240,0.06066133578618368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,12288,0.06002133422427707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,8192,0.042549331982930504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,10240,0.049604445695877075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,7168,0.037676443656285606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,8192,0.044870220952563815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,6144,0.03283199999067519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,7168,0.04160533348719279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,6144,0.03604711095492045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,5120,0.02823022339079115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,4096,0.023052444060643513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,5120,0.03125066558519999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,3584,0.02087733315096961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,4096,0.02723199956946903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,12288,0.06994577911165026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,3072,0.018232888645595975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,3584,0.025937777426507738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,2560,0.015653333730167814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,2560,0.02016888890001509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,3072,0.023257777094841003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,2048,0.012958221965365939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,1536,0.01070755554570092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,2048,0.019225777851210702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,1024,0.00815466625822915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,1536,0.017602667212486267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,768,0.007053333024183909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,1024,0.014775110615624322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,512,0.004943999979231092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,768,0.014109333356221518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,256,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,512,0.012725333372751871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,128,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,256,0.012392889294359418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,64,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,128,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,16384,32,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,64,0.011715555356608497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,16384,32,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,65536,0.24484711223178438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,65536,0.15557777881622314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,51200,0.12035111586252849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,16384,0.06983733177185059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,16384,0.04830577638414171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,12288,0.050979554653167725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,12288,0.04734488990571764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,51200,0.19307733906639946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,10240,0.04382222228580051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,8192,0.036815110180113055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,10240,0.04385333259900411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,7168,0.03259199857711792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,8192,0.04155022237035964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,6144,0.028552889823913574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,5120,0.024109333753585815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,7168,0.036751111348470054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,4096,0.019768888751665752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,6144,0.03405866689152188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,5120,0.026405332816971674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,4096,0.024711110525661047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,3584,0.01787911189926995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,3072,0.015464888678656684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,2560,0.013572444518407186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,3584,0.02369155486424764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,2048,0.011711999773979187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,2560,0.01904533306757609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,1536,0.009697777529557547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,2048,0.01751377847459581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,3072,0.0221742226017846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,1024,0.007697777615653143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,768,0.006574222197135289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,1536,0.016417778200573392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,512,0.004540444248252445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,1024,0.014755555325084262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,768,0.013422222601042854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,256,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,128,0.0034213333080212274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,512,0.01239111108912362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,64,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,256,0.012138666378127204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,12288,32,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,64,0.010847110715177325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,32,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,12288,128,0.011855111353927188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,65536,0.20594222015804717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,51200,0.1595404413011339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,65536,0.15080622831980386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,16384,0.06369600031110975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,12288,0.04533866710133023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,51200,0.11595822705162896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,16384,0.046671112378438316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,10240,0.038554665115144514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,8192,0.031189332405726116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,12288,0.04553866717550489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,10240,0.03849333193567064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,7168,0.028910222980711196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,8192,0.03202666507826911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,6144,0.02606133288807339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,5120,0.022619555393854778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,7168,0.030717333157857258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,6144,0.026320000489552815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,4096,0.01796888808409373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,3584,0.016232000456915963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,4096,0.02171733313136631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,3072,0.014504000544548035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,5120,0.025021332833502028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,3584,0.021047110358874004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,2560,0.0129120002190272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,2048,0.010972444381978778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,3072,0.019462222854296368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,2560,0.01758311192194621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,1536,0.009378666679064432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,1024,0.007673777639865875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,2048,0.01718844473361969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,1536,0.01516533394654592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,768,0.0053102220926019884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,1024,0.012845333251688214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,512,0.004610666798220741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,768,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,256,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,512,0.011782222323947482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,128,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,256,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,64,0.003170666595300039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,128,0.010779555473062726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,10240,32,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,64,0.0114720000161065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,10240,32,0.010354667074150508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,65536,0.16224088933732775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,65536,0.10032000144322713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,51200,0.12876089413960776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,16384,0.051220443513658315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,51200,0.07791378100713094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,12288,0.03256888853179084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,16384,0.03490577803717719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,10240,0.02803022331661648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,12288,0.03400266501638625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,8192,0.022967999180157978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,10240,0.03152622116936578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,7168,0.020784000555674236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,8192,0.025752888785468206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,6144,0.01832711034350925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,7168,0.02476799984773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,5120,0.016152888536453247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,6144,0.022515555222829182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,4096,0.013106666505336761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,5120,0.02055644492308299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,3584,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,4096,0.01881511178281572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,3072,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,3584,0.018588445252842374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,2560,0.009873777627944946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,3072,0.017475555340449016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,2048,0.008619555168681675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,2560,0.01584533353646596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,1536,0.007351111206743453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,2048,0.014484445254007975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,1024,0.005387555393907759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,1536,0.013691555294725629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,768,0.004208000169859992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,1024,0.012406222522258759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,512,0.0034942221310403612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,768,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,256,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,512,0.011425777441925473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,128,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,256,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,64,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,128,0.010417777631017896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,8192,32,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,64,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,65536,0.15165689256456164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,65536,0.09150577916039361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,51200,0.11408978038363987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,51200,0.07177155547671847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,16384,0.042267554336123996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,16384,0.032878223392698504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,8192,32,0.010063110954231685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,12288,0.033188445700539485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,10240,0.02851466668976678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,12288,0.032941334777408175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,8192,0.024102222588327196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,10240,0.029827555020650227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,7168,0.0218577782313029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,8192,0.025642666551801894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,7168,0.024685333172480266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,6144,0.02195377813445197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,5120,0.014927999840842353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,5120,0.020283554991086323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,4096,0.012230222423871359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,4096,0.018789332773950364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,3584,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,3584,0.018495111001862418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,3072,0.01037333326207267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,6144,0.01677688956260681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,2560,0.009164444274372524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,3072,0.017287110288937885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,2048,0.008104888929261101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,2560,0.015184000134468079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,1536,0.006979555719428592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,2048,0.01391377713945177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,1024,0.004555555681387584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,1536,0.01312888910373052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,768,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,1024,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,512,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,768,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,256,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,512,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,128,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,256,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,64,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,128,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,7168,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,64,0.010665777656767102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,7168,32,0.010694222317801582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,65536,0.12861155139075384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,51200,0.09526399771372478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,65536,0.08281777964697944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,51200,0.06542222367392646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,16384,0.03673422336578369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,16384,0.030721777015262183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,12288,0.02786933382352193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,12288,0.03313510947757297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,10240,0.02458488941192627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,10240,0.02743822170628442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,8192,0.020901333954599168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,8192,0.025369778275489807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,7168,0.01882488860024346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,7168,0.02422399984465705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,6144,0.01699911057949066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,6144,0.02131822208563487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,5120,0.015117333994971381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,5120,0.01955111159218682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,4096,0.011570666399266986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,4096,0.01808888879087236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,3584,0.010694222317801582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,3584,0.017851556340853374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,3072,0.009504000345865885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,3072,0.016781333420011733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,2560,0.008630221916569604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,2560,0.014428445034556918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,2048,0.007448888487286038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,2048,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,1536,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,1536,0.013065777719020844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,1024,0.004182222402758068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,1024,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,768,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,768,0.012061333490742577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,512,0.0032479998966058097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,512,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,256,0.0031786666562159858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,128,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,128,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,64,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,256,0.010986666712496014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,64,0.011152000062995486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,6144,32,0.002999111182159848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,65536,0.10623644457923041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,6144,32,0.010747555229398938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,51200,0.08677777979109023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,51200,0.05991466840108236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,65536,0.07441866397857666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,16384,0.03445511062939962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,12288,0.02574488851759169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,12288,0.030591110388437908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,16384,0.02879555688963996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,10240,0.0221404449807273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,8192,0.019285332825448778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,8192,0.024311999479929607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,10240,0.025998221503363714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,7168,0.02291733357641432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,6144,0.01533422205183241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,6144,0.020454222957293194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,5120,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,7168,0.01732711162832048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,4096,0.010568888651000129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,5120,0.018615111708641052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,3584,0.009630222287442949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,4096,0.017571555243598092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,3072,0.008709333009190029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,3584,0.016952888833151925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,2560,0.007996444072988298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,3072,0.01592177814907498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,2048,0.0069679998689227635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,2560,0.013818666338920593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,2048,0.01329066687160068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,1536,0.00509333320789867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,1024,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,1536,0.012798222402731577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,768,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,1024,0.011895111037625207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,512,0.003616888903909259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,768,0.012160888976520963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,256,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,512,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,256,0.011477332976129321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,128,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,64,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,128,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,5120,32,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,64,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,5120,32,0.01037599974208408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,65536,0.08650488985909356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,51200,0.06668000088797675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,65536,0.06843199994828966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,16384,0.027520888381534155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,51200,0.05515644285413954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,12288,0.02102577851878272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,16384,0.025459556115998164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,10240,0.018449778358141582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,12288,0.02956355611483256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,8192,0.015549333559142219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,10240,0.02349333299530877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,7168,0.01368444412946701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,8192,0.023743111226293776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,6144,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,7168,0.021057777934604224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,6144,0.018888889087571036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,5120,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,5120,0.01772177716096242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,4096,0.009370666411187913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,4096,0.016539555456903245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,3584,0.00886222223440806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,3584,0.0151582227812873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,3072,0.008034666379292807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,3072,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,2560,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,2560,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,2048,0.005704888867007361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,2048,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,1536,0.004560889055331548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,1536,0.01276088919904497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,1024,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,1024,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,768,0.003549333247873518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,768,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,512,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,512,0.011695110963450538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,256,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,256,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,128,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,64,0.002546666603949335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,64,0.010764444039927589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,4096,32,0.002502222235004107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,4096,32,0.01036000003417333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,65536,0.07589599821302626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,65536,0.0657600031958686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,51200,0.05875644418928358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,16384,0.024199111594094172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,51200,0.05327199896176656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,12288,0.018400889303949144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,16384,0.02474044428931342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,12288,0.02568000058333079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,10240,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,8192,0.013939556148317126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,10240,0.02272355556488037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,7168,0.013714666995737286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,8192,0.022424888279702928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,6144,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,7168,0.02017333275742001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,5120,0.010871999793582492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,6144,0.019559999306996662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,4096,0.009118222528033787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,5120,0.01698666645420922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,3584,0.008741333252853816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,4096,0.015847111741701763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,3072,0.007833778030342525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,3584,0.014546665880415173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,2560,0.006909333169460297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,2560,0.012784000072214337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,2048,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,3072,0.014927999840842353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,2048,0.012895999683274163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,1536,0.012465777496496836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,1024,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,1024,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,768,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,768,0.01183733344078064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,512,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,256,0.0033724444607893624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,512,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,1536,0.004560889055331548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,128,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,64,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,256,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3584,32,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,128,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,64,0.011138666835096149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,65536,0.06301955382029216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3584,32,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,51200,0.05050311154789395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,65536,0.06262311008241442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,16384,0.02130844526820713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,12288,0.01660977800687154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,51200,0.05085244443681505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,10240,0.016947555873129103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,16384,0.023805333508385554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,8192,0.014803555276658801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,12288,0.0243848893377516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,10240,0.022376888328128394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,7168,0.012440889245933957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,6144,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,8192,0.020288000504175823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,5120,0.010086221827401055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,6144,0.018256889449225534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,5120,0.015504888362354703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,4096,0.008809777597586313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,3584,0.008200888832410177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,4096,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,3072,0.00703022215101454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,3584,0.013824000126785703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,2560,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,3072,0.013560000393125745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,2048,0.00527022240890397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,2560,0.013171555267439948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,2048,0.012445333103338877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,1536,0.00423555572827657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,7168,0.019909333851602342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,1024,0.003948444293604957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,1536,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,768,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,1024,0.01146666705608368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,512,0.0032640000184377036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,768,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,256,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,512,0.011119999819331698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,128,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,256,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,64,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,128,0.010445333189434476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,3072,32,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,64,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,65536,0.05560355716281467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,3072,32,0.009759999811649323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,51200,0.04551377892494202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,65536,0.060237334834204785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,16384,0.0196079992585712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,51200,0.04897600081231859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,12288,0.014831999937693277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,16384,0.023035556077957153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,12288,0.022614222433831956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,10240,0.013893332746293811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,8192,0.011864888999197217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,10240,0.020922667450375024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,7168,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,8192,0.019653333557976615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,6144,0.01055466632048289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,7168,0.018917333748605516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,5120,0.009672889278994666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,6144,0.016536888149049547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,4096,0.008340444829728868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,5120,0.014558222558763293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,3584,0.007831999825106727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,4096,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,3072,0.006695999867386288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,3584,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,2560,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,3072,0.013235555754767524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,2048,0.004926222066084544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,2560,0.012138666378127204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,1536,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,2048,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,1024,0.004002666721741359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,1536,0.011794666449228922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,768,0.0037173334923055437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,1024,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,512,0.003195555466744635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,768,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,256,0.0031831111345026228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,512,0.01144177797767851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,128,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,256,0.01074044406414032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,64,0.002564444517095884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,128,0.011113777756690979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2560,32,0.0025271111064487034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,64,0.010370666782061258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,65536,0.04558755622969734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2560,32,0.010419555836253695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,51200,0.03781777620315552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,65536,0.057836446497175426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,16384,0.01518311103185018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,51200,0.04779111014472114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,12288,0.01292266696691513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,16384,0.022346667117542688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,10240,0.013198222551080914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,12288,0.021195555726687115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,8192,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,10240,0.019984000258975558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,7168,0.011045333411958484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,8192,0.01787911189926995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,6144,0.010113777385817634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,7168,0.01648533344268799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,5120,0.009714666340086196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,6144,0.014801777071423002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,4096,0.00795288880666097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,5120,0.013791110780504016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,3584,0.0069653333889113525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,4096,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,3072,0.005912888795137405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,3584,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,2560,0.005631111148330901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,3072,0.013110222087966071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,2048,0.004886222382386525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,2560,0.012127111355463663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,1536,0.004587555511130227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,2048,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,1024,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,1536,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,768,0.003735110991530948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,1024,0.011356444822417365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,512,0.0032106666929192017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,768,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,256,0.0032382222513357797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,512,0.011367110742463006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,128,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,256,0.010727110836240979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,128,0.010388444695207808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,64,0.0026613333159022858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,64,0.010307555397351583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,2048,32,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,2048,32,0.009724444813198512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,65536,0.035754667388068304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,65536,0.056040002240075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,51200,0.029017776250839233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,51200,0.04592622319857279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,16384,0.014730667074521383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,16384,0.02200355629126231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,12288,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,12288,0.020073778099483915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,10240,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,10240,0.018219555417696636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,8192,0.010055111514197456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,8192,0.016529776983790927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,7168,0.010222222242090438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,7168,0.015187554889255099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,6144,0.009396444592210982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,6144,0.014767110347747803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,5120,0.00833155545923445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,5120,0.013817778064144982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,4096,0.007383999725182851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,4096,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,3584,0.006660444455014334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,3584,0.012527111503813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,3072,0.005953777581453323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,3072,0.012755555411179861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,2560,0.005596444424655702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,2560,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,2048,0.004917333523432414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,2048,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,1536,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,1536,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,1024,0.0038613333470291565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,1024,0.011110222174061669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,768,0.003568888952334722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,768,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,512,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,512,0.011018666956159802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,256,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,256,0.0107342220014996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,128,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,128,0.010432889064153036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,64,0.002830222249031067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,64,0.010271111296282874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1536,32,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1536,32,0.009662222531106737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,65536,0.025239111648665533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,65536,0.053766220808029175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,51200,0.02130222154988183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,51200,0.043360888957977295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,16384,0.011149333582984077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,16384,0.021387555532985266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,12288,0.010864888628323873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,12288,0.018209778600268893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,10240,0.010041777458455827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,10240,0.017151999804708693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,8192,0.00871377769443724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,8192,0.015782222151756287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,7168,0.009160888691743216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,7168,0.01516177753607432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,6144,0.00870222184393141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,6144,0.014483556151390076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,5120,0.00776800016562144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,5120,0.013847110999955071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,4096,0.007349333001507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,3584,0.006704888823959563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,4096,0.013409777647919126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,3072,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,3584,0.013099555340078143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,2560,0.00555288874440723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,3072,0.012724444270133972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,2048,0.006610666712125142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,2560,0.012447111308574677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,2048,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,1536,0.0053093334039052325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,1536,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,768,0.0041991112132867174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,1024,0.011131555669837527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,512,0.00361333332127995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,768,0.011740444435013665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,256,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,128,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,512,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,64,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,128,0.010420444111029306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,1024,32,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,64,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,256,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,1024,32,0.009735999835862054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,65536,0.023205333285861548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,51200,0.019330667124854196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,16384,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,65536,0.052539553907182485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,51200,0.04318222072389391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,16384,0.02036711076895396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,12288,0.010418666733635796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,10240,0.009530666801664565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,12288,0.01811644434928894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,8192,0.008316444853941599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,10240,0.016916443904240925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,7168,0.008707555631796518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,8192,0.015568888849682279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,6144,0.00813244448767768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,7168,0.015365334020720588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,5120,0.007352888584136963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,6144,0.01426755554146237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,4096,0.007044444481531779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,5120,0.014174222946166992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,3584,0.006605333338181178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,4096,0.01311466677321328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,3072,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,3584,0.013117333253224691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,2560,0.005584888988071018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,3072,0.012776888906955719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,2048,0.005009777843952179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,2560,0.012770666844315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,1536,0.004931555440028508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,2048,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,1024,0.003940444439649582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,1536,0.012175111307038201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,768,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,1024,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,512,0.0033031110134389666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,512,0.010855110983053843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,768,0.012014221813943652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,256,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,128,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,256,0.011819555527634092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,128,0.010429333481523724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,64,0.002578666640652551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,768,32,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,64,0.0107351111041175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,65536,0.018555555078718398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,768,32,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,51200,0.016953777935769822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,65536,0.052058666944503784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,16384,0.009752888646390703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,51200,0.04314577910635206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,12288,0.009674666656388177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,16384,0.01999022232161628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,10240,0.008737777670224508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,12288,0.018239999810854595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,10240,0.016583111551072862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,8192,0.008041777544551426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,7168,0.008763555851247575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,8192,0.015494222442309061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,6144,0.008031110796663497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,7168,0.015185778339703878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,5120,0.007682667010360294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,6144,0.014336888988812765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,4096,0.006956444846259222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,5120,0.014230221509933472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,3584,0.006673777682913675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,4096,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,3072,0.006695999867386288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,3584,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,2560,0.006395555618736479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,3072,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,2048,0.00565244464410676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,2560,0.012777778009573618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,1536,0.004585777719815572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,2048,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,1024,0.003654222107595868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,1536,0.012468444804350534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,1024,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,768,0.0036071112586392295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,512,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,768,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,256,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,512,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,128,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,256,0.010813333094120026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,64,0.0025946667624844448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,128,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,512,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,64,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,65536,0.016237333416938782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,512,32,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,51200,0.01568800045384301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,16384,0.009168888959619734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,65536,0.0491226679748959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,12288,0.007662222617202335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,51200,0.040327111879984535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,10240,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,16384,0.02035999960369534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,8192,0.006826666494210561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,12288,0.017992888887723286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,7168,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,10240,0.016872000363137987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,6144,0.007289778027269576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,8192,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,5120,0.006707555717892117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,7168,0.015076445208655464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,4096,0.0058986664646201665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,6144,0.014615999327765571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,5120,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,3584,0.0062977779242727495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,4096,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,3072,0.0052577778697013855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,3584,0.01310755560795466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,2560,0.006082666830884085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,3072,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,2560,0.012088889049159156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,2048,0.005255110975768831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,1536,0.004584889031118817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,2048,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,1024,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,1536,0.011527111132939657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,768,0.003587555554178026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,1024,0.011069333387745751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,512,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,768,0.01106844428512785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,256,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,512,0.011111111276679568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,128,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,256,0.010570666856235929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,64,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,128,0.011134222149848938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,256,32,0.0027751111321979095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,64,0.010775999890433418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,65536,0.014127999544143677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,256,32,0.010096000300513374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,51200,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,65536,0.04736888739797804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,16384,0.007671111159854465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,51200,0.03927377859751383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,12288,0.007303111255168915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,16384,0.01998222205373976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,10240,0.0069617778062820435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,12288,0.018039110634062026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,8192,0.006287999865081575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,10240,0.01647822227742937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,8192,0.015856888559129503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,7168,0.006038222048017714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,6144,0.005580444302823808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,7168,0.015139554937680563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,5120,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,6144,0.014135110709402295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,4096,0.005149333427349727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,5120,0.013786666923099093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,3584,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,4096,0.013087111214796701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,3072,0.0052382221652401825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,3584,0.013085333009560903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,2560,0.005928888916969299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,3072,0.012660444610648684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,2048,0.0052871112194326185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,2560,0.012299555871221753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,1536,0.004201777693298128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,2048,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,1536,0.01182044463025199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,1024,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,768,0.0034835553831524322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,1024,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,512,0.0032026666320032547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,768,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,256,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,512,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,128,0.0025111111915773815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,256,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,64,0.002504000026318762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,128,0.01107022249036365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,128,32,0.00250133333934678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,64,0.010693333215183683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,65536,0.0107022225856781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,128,32,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,51200,0.010072888599501716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,65536,0.04698044392797682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,16384,0.00665511108107037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,51200,0.039417776796552874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,12288,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,16384,0.019924443629052904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,12288,0.017863111363516915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,10240,0.006239999913507038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,8192,0.006015111174848344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,10240,0.016885333591037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,7168,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,8192,0.015479111009173922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,6144,0.00556355549229516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,7168,0.015214223000738354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,5120,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,6144,0.014123555686738757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,4096,0.005241777747869492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,5120,0.014185777968830533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,3584,0.005952000204059813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,4096,0.01313333296113544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,3072,0.005557333429654439
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,3584,0.01314399970902337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,2560,0.005913777897755305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,3072,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,2560,0.012308444413873883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,2048,0.0052391112678580815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,1536,0.004242666479614046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,2048,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,1024,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,1536,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,1024,0.01109777804878023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,768,0.0033653332955307434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,768,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,512,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,512,0.011015999648306104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,256,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,256,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,128,0.002508444504605399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,128,0.010445333189434476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,64,0.0024986666523747975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,64,0.010699555277824402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,64,32,0.002524444419476721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,64,32,0.010428444378905825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,65536,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,65536,0.04731733269161648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,51200,0.010192888478438059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,51200,0.03906666570239597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,16384,0.0069013333155049225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,12288,0.006333333336644703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,16384,0.019976889093716938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,10240,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,12288,0.017714665995703805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,8192,0.005903999838564131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,10240,0.01681155628628201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,7168,0.005922666854328579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,8192,0.015583111180199517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,6144,0.005610666755172942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,7168,0.014911111858155994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,5120,0.005252444495757421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,6144,0.01443822185198466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,4096,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,5120,0.013822221921549903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,4096,0.013201778133710226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,3584,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,3072,0.00527022240890397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,3584,0.012429333395428128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,2560,0.00555288874440723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,3072,0.012681778106424542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,2048,0.0052133335007561584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,2560,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,1536,0.0042257776690853965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,2048,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,1024,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,1536,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,768,0.0034862222770849862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,1024,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,512,0.0031439999325407874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,768,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,256,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,256,0.010413332945770687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,128,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,512,0.011164444188276926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,64,0.002522666628162066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,128,0.010424888796276517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,10,32,32,0.00282222218811512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,64,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,10,32,32,0.009524444739023844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,51200,0.4937742021348741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,51200,0.9508203930324979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,16384,0.33745066324869794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,12288,0.24096979035271537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,16384,0.1671137809753418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,10240,0.18912443849775526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,12288,0.19123555554283989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,8192,0.15224711100260416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,10240,0.12726133399539524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,7168,0.1336453358332316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,8192,0.11463733514149983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,6144,0.11519111527336968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,6144,0.09950044420030381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,7168,0.10456711053848267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,5120,0.09784089194403754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,4096,0.0775760014851888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,3584,0.0687733358807034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,5120,0.08436089091830784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,3072,0.05976444482803345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,4096,0.06935999790827434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,2560,0.05058133271005419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,3584,0.05927199787563748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,2048,0.041259557008743286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,2560,0.04830844534767998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,1536,0.03222577770551046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,2048,0.04054488738377889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,1024,0.023027555810080633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,1536,0.033032000064849854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,768,0.018526222970750596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,1024,0.02492888934082455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,512,0.013152889079517789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,768,0.022474666436513264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,256,0.00903200027015474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,512,0.018696889281272888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,128,0.006278222219811545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,3072,0.05288266804483202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,64,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,256,0.01588000026014116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,65536,32,0.005819555371999741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,128,0.01311911145846049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,64,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,65536,32,0.015829333000712924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,65536,0.5198888778686523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,65536,1.0332809024386935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,51200,0.8004693455166287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,51200,0.39918843905131024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,16384,0.2745600011613634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,16384,0.13679111003875732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,12288,0.1988133324517144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,12288,0.1478497717115614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,10240,0.16546844111548528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,8192,0.13731466399298772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,10240,0.10468355814615886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,7168,0.11969510714213054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,8192,0.09331466754277547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,6144,0.10374310943815444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,6144,0.08300355407926771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,7168,0.08573066525989109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,5120,0.08534400330649482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,4096,0.07115288575490315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,5120,0.06365422407786052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,3584,0.06253155734803942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,3584,0.05208711160553826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,4096,0.057822220855289035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,3072,0.05418577790260315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,2560,0.04543022314707438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,3072,0.04653155472543505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,2048,0.0381991101635827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,2560,0.04082489013671875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,1536,0.029571556382709082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,2048,0.03503288825352987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,1024,0.021697777840826247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,1536,0.029679109652837116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,1024,0.022476444641749065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,768,0.0163137780295478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,512,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,768,0.020608888732062448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,256,0.007694222033023834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,512,0.01722400055991279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,128,0.004501333253251182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,256,0.01347555551264021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,64,0.003565333369705412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,128,0.012469333079126147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,51200,32,0.003579555700222651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,32,0.014143110977278816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,65536,0.3295351134406196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,65536,0.1804577774471707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,51200,64,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,51200,0.2592426670922173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,51200,0.13913244671291775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,16384,0.09524266587363349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,16384,0.05407022105322944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,12288,0.06954666641023424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,10240,0.060067554314931236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,12288,0.05989333656099108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,8192,0.042862223254309766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,10240,0.04933866527345446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,7168,0.037772443559434675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,8192,0.04287733303176033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,6144,0.03317244516478644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,6144,0.03720088799794515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,5120,0.028532445430755615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,7168,0.040263109736972384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,4096,0.02313422163327535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,3584,0.020642666353119742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,5120,0.032101333141326904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,3072,0.018197332819302876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,4096,0.02900800108909607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,2560,0.015853333804342482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,3584,0.025545777546034917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,3072,0.02296977738539378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,2048,0.012729778057999082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,1536,0.010704889065689512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,2048,0.018795554836591084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,1024,0.008172444171375698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,2560,0.021831999222437542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,768,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,1024,0.01477866702609592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,512,0.00489155575633049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,512,0.01276177747382058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,256,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,768,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,128,0.003129777808984121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,256,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,64,0.0031617778456873367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,64,0.010716444088353051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,128,0.011148444480366178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,1536,0.017895999881956313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,16384,32,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,65536,0.24487198723687065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,51200,0.19300356176164415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,65536,0.15490755769941542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,16384,0.0701191094186571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,51200,0.12029777632819282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,12288,0.05028088887532552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,16384,32,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,16384,0.0483822226524353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,10240,0.043918222188949585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,8192,0.0366266667842865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,12288,0.051129778226216636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,10240,0.042556444803873696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,7168,0.03243111239539252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,8192,0.03724799884690179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,6144,0.028725332683987085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,7168,0.03448977735307481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,5120,0.023815999428431194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,6144,0.033301333586374916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,4096,0.01961955593691932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,5120,0.026762665973769292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,4096,0.02427022159099579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,3584,0.01755466726091173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,3072,0.015789333317014907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,3584,0.022985777921146814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,2560,0.01369511087735494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,2560,0.019564444820086162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,3072,0.0215013325214386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,2048,0.011936888926559024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,1536,0.009711110757456886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,2048,0.017975111802419026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,1536,0.016591111818949383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,1024,0.007804444266690149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,768,0.006716444260544247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,1024,0.013693333499961428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,512,0.004561777744028303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,768,0.013187555803192986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,256,0.00360622215602133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,512,0.012420444852775998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,128,0.003536888708670934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,256,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,64,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,128,0.010430222584141625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,12288,32,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,64,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,12288,32,0.010431110858917236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,65536,0.20515199502309164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,65536,0.14684089024861655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,51200,0.15986666414472792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,16384,0.06286488638983832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,51200,0.11543467309739854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,12288,0.04544622368282742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,12288,0.042903999487559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,16384,0.04689244429270426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,10240,0.03849244448873732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,8192,0.0311173333062066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,10240,0.03773599863052368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,7168,0.028941333293914795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,8192,0.032511999209721885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,6144,0.025962667332755193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,7168,0.029898666673236426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,6144,0.026927999324268762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,5120,0.022464000516467627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,4096,0.017768000562985737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,5120,0.024335111180941265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,3584,0.015999999311235216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,4096,0.02235466738541921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,3072,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,3584,0.020582222276263766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,2560,0.012714666624863943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,3072,0.018983999888102215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,2560,0.01795288920402527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,2048,0.0162124451663759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,1536,0.009350222018029954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,1536,0.015217777755525378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,1024,0.007583111524581909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,1024,0.012891555825869242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,768,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,768,0.012824888858530255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,2048,0.010975110861990186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,512,0.004969777746333016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,256,0.00378577783703804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,128,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,512,0.01204355557759603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,64,0.003133333391613431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,256,0.011794666449228922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,128,0.010689777632554373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,64,0.010633777413103314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,10240,32,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,65536,0.16291466024186876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,10240,32,0.011115555961926779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,65536,0.10078489118152195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,51200,0.12836799356672499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,16384,0.05211911267704434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,51200,0.07780800263086955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,12288,0.032646222247017756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,10240,0.02814399864938524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,16384,0.034733331865734525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,12288,0.03436711099412706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,8192,0.022915555371178523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,7168,0.020618667205174763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,10240,0.030949334303538006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,8192,0.027099554737408955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,7168,0.025832000705930922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,6144,0.01847822301917606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,5120,0.01593866613176134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,4096,0.013217777841620974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,6144,0.022455111145973206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,3584,0.012040889097584618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,5120,0.02096533278624217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,4096,0.018886221779717337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,3072,0.010882666541470421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,2560,0.009647111097971598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,3072,0.016911110944218107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,3584,0.018256889449225534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,2048,0.00832799987660514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,2560,0.016846223009957206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,1536,0.007188444336255391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,2048,0.014857777290874057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,1024,0.004954666727119022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,1536,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,768,0.003933333274390963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,1024,0.012440889245933957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,512,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,768,0.012930666406949362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,256,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,512,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,128,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,256,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,64,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,128,0.010791111323568555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,8192,32,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,64,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,8192,32,0.011023999916182624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,65536,0.15360444121890598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,51200,0.11391288704342312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,65536,0.09070222245322333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,16384,0.04249422086609734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,51200,0.07159644365310669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,16384,0.0328693323665195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,12288,0.0331911113527086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,10240,0.028914666838116113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,12288,0.03439822130733066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,10240,0.0299217767185635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,8192,0.02407377792729272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,7168,0.02176977859603034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,8192,0.026142223013771906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,6144,0.016940444707870483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,7168,0.024448888169394598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,5120,0.014836443795098199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,6144,0.022426666484938726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,5120,0.019741333193249173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,4096,0.012349333200189801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,3584,0.011375111010339526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,4096,0.01923022170861562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,3072,0.010035555395815108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,3584,0.018876444962289598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,2560,0.009139555195967356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,3072,0.016927111479971144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,2560,0.0162764439980189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,2048,0.007994666695594788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,1536,0.006630222416586346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,2048,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,1024,0.0042453333735466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,1536,0.013452444639470843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,768,0.0035822221802340615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,1024,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,768,0.012145777543385824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,512,0.003440888805521859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,256,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,512,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,256,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,128,0.0029680000411139596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,64,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,128,0.011053333679835001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,64,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,7168,32,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,65536,0.1278862158457438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,7168,32,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,51200,0.09489155477947658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,65536,0.08189066913392809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,16384,0.0365226666132609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,51200,0.06532622045940824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,12288,0.02793688906563653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,16384,0.030745777818891738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,10240,0.024526221884621516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,12288,0.03157777918709649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,8192,0.020593777298927307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,10240,0.026520889666345384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,8192,0.026527111728986103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,7168,0.018935110833909776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,6144,0.016977777083714802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,7168,0.02312177750799391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,5120,0.014821334017647637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,6144,0.02130399975511763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,4096,0.011384888655609555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,5120,0.019875556230545044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,3584,0.010607110957304636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,4096,0.01886399918132358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,3072,0.009791110952695211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,3584,0.01794577803876665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,2560,0.008671999805503422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,3072,0.016916443904240925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,2048,0.0074888889988263445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,2560,0.015807999504937064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,1536,0.006605333338181178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,2048,0.013829333086808523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,1024,0.004317333300908406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,1536,0.01314311143424776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,768,0.003768889026509391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,1024,0.012599111431174807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,512,0.0033084443873829315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,768,0.012119111087587146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,256,0.0032097777972618737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,512,0.012151111331250934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,256,0.011677333050303988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,128,0.003112000102798144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,128,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,64,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,64,0.011507555842399597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,6144,32,0.0029164445069101122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,6144,32,0.011176000038782755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,65536,0.10625688897238837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,65536,0.07459821965959337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,51200,0.08660711182488336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,16384,0.03480444351832072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,51200,0.0599457754029168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,16384,0.028513779242833454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,12288,0.025758221745491028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,10240,0.021864889396561518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,12288,0.0288426677385966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,8192,0.019294222195943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,10240,0.02475555572244856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,7168,0.01723288827472263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,8192,0.024724443753560383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,6144,0.015542222393883599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,7168,0.02295555505487654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,5120,0.013617777989970313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,6144,0.020012444920010038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,4096,0.01036533299419615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,5120,0.01852888862291972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,3584,0.009687999884287516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,4096,0.01810933318403032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,3072,0.008758222063382467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,3584,0.017495110630989075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,2560,0.00814400033818351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,3072,0.015178667174445258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,2048,0.007215110792054071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,2560,0.014856888188256158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,1536,0.005251555393139522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,2048,0.01348444405529234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,1024,0.003910222401221593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,1536,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,768,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,1024,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,512,0.0035217776894569397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,768,0.012464889221721225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,256,0.002982222164670626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,512,0.011767999993430244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,128,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,256,0.011718221836619906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,64,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,128,0.010856889188289642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,5120,32,0.002524444419476721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,64,0.01148444414138794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,65536,0.0862542192141215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,65536,0.06798399819268121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,51200,0.06655644708209567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,51200,0.055051555236180626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,16384,0.028168890211317275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,16384,0.025158221522967022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,12288,0.020815999971495736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,5120,32,0.01034755590889189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,10240,0.018535999788178336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,12288,0.02723466687732273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,8192,0.015469332536061605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,10240,0.022679110368092854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,7168,0.013983999689420065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,8192,0.022895110978020564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,6144,0.012543111211723752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,7168,0.020199111766285367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,5120,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,6144,0.019039111004935372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,4096,0.009658666948477427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,5120,0.017200888858901132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,3584,0.0088746663596895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,4096,0.01698844465944502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,3584,0.015607999430762397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,3072,0.00812355594502555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,2560,0.007319110963079665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,3072,0.014148443937301636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,2048,0.006005333529578314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,2560,0.014448889427714877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,1536,0.004535111288229625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,2048,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,1024,0.0038906666967603895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,1536,0.013064888616402945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,768,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,1024,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,512,0.0032479998966058097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,768,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,256,0.0031635556370019913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,512,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,256,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,64,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,128,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,4096,32,0.0025742221623659134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,32,0.01036888857682546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,4096,64,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,65536,0.07528711027569242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,51200,0.0584826668103536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,65536,0.06518310970730253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,51200,0.05278933379385206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,16384,0.023807111713621352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,16384,0.024438222249348957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,12288,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,10240,0.01637066735161675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,12288,0.02643644478585985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,10240,0.022469333476490442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,8192,0.014124444789356656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,7168,0.01347555551264021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,8192,0.02241244415442149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,7168,0.021341333786646526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,6144,0.01182044463025199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,5120,0.010812444819344414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,6144,0.019282667173279654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,4096,0.009097778134875828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,5120,0.01737600068251292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,3584,0.008522666990756989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,4096,0.014912888407707214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,3072,0.007988444632954067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,3584,0.015498666299713982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,2560,0.006728000111050076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,3072,0.013776000175211163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,2048,0.005277333160241445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,2560,0.014175999495718213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,1536,0.004527111020353106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,2048,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,1024,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,1536,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,768,0.003847111016511917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,1024,0.011483555866612328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,512,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,768,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,512,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,256,0.010688888529936472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,128,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,128,0.01033866653839747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,64,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,64,0.010404444403118556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,32,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3584,32,0.01000088867213991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3584,256,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,65536,0.06272533204820421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,51200,0.05049688948525322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,65536,0.062450667222340904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,16384,0.021207110749350652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,16384,0.023553777072164748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,12288,0.016533333394262526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,12288,0.02439466615517934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,10240,0.017288888494173687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,10240,0.021070222059885662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,8192,0.01457333399189843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,51200,0.05066577924622429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,8192,0.020484444167878892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,7168,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,6144,0.011047999891969891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,7168,0.019552888141738046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,6144,0.018591110905011494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,5120,0.010038221875826517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,4096,0.008791999684439765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,5120,0.016548444827397663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,3584,0.00816177742348777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,4096,0.01477599971824222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,3072,0.0075697774688402815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,3584,0.014412444498803882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,2560,0.0063288890653186375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,3072,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,2048,0.005393777870469623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,2560,0.013087111214796701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,1536,0.00464088883664873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,2048,0.012464000119103326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,1024,0.0038702223036024305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,1536,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,768,0.0035911111368073356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,1024,0.012083555261294046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,768,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,512,0.0032888888898822996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,256,0.003118222165438864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,512,0.011344888971911537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,128,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,256,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,64,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,128,0.010264889233642155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,3072,32,0.0028328889360030494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,64,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,65536,0.05552533268928528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,3072,32,0.009761778016885122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,51200,0.0454711119333903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,65536,0.059506667984856494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,16384,0.01954844428433312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,51200,0.049040890402264066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,12288,0.014816888504558139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,16384,0.02312177750799391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,10240,0.013827555709415011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,12288,0.023530667026837666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,8192,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,10240,0.020428443948427837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,8192,0.018944000204404194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,6144,0.010781333678298526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,7168,0.01864888932969835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,6144,0.016951110627916124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,5120,0.014867555763986377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,7168,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,4096,0.00832622249921163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,5120,0.009690666364298927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,3584,0.007661333514584436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,4096,0.014160000615649752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,3072,0.006717333363162146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,3584,0.01423733267519209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,2560,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,2560,0.01350488927629259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,2048,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,2048,0.012464000119103326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,1536,0.004574222283230888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,1536,0.012497777740160624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,1024,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,1024,0.011805333197116852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,768,0.0036159998012913596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,3072,0.013168888787428537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,512,0.0033173333439562055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,768,0.012216889195972018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,256,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,512,0.01145066652033064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,256,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,128,0.01036977767944336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2560,32,0.0026231110095977783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,64,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,65536,0.04589066571659512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2560,32,0.010425777898894416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,51200,0.03762044509251913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,65536,0.05767644113964505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,16384,0.015048889650238885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,51200,0.04752177662319607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,12288,0.012841777669058906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,16384,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,10240,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,12288,0.021957332889238994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,8192,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,10240,0.019454222586419847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,7168,0.01128088848458396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,8192,0.01852799952030182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,6144,0.010404444403118556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,7168,0.016275554895401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,5120,0.00942311104800966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,6144,0.015217777755525378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,4096,0.008049777812427944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,5120,0.014491554763582019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,3584,0.007064888874689738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,4096,0.013506666653686099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,3072,0.006267555471923616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,3584,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,2560,0.00565422202150027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,3072,0.012888888518015543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,2048,0.005286222116814719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,2560,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,1536,0.004272888931963178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,2048,0.012146666646003723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,1024,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,1536,0.01237422227859497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,768,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,1024,0.01145244472556644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,512,0.00351200004418691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,768,0.012146666646003723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,256,0.002983111060327954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,512,0.011159111228254108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,128,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,256,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,64,0.00254844439526399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,128,0.010438222024175854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,2048,32,0.0028719999310043124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,64,0.01148977792925305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,65536,0.0358257790406545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,2048,32,0.010503110786279043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,51200,0.02901866700914171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,65536,0.056314667065938316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,16384,0.014892444014549255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,51200,0.04561333192719353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,12288,0.011932444241311816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,16384,0.022356443934970435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,10240,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,12288,0.020711110697852243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,8192,0.010100444157918295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,10240,0.018304889400800068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,7168,0.010290666586822933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,8192,0.016162667009565566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,6144,0.009538667069541084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,7168,0.015511999527613321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,5120,0.008424888882372115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,6144,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,4096,0.007321777443091075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,5120,0.014478221535682678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,3584,0.006726222319735422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,4096,0.013176889055305056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,3072,0.005986666513813867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,2560,0.005738666488064661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,3584,0.013864888913101621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,3072,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,2048,0.005012444323963589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,1536,0.004495999879307217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,2560,0.01242844429281023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,2048,0.012106666962305704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,1024,0.0038933331767717996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,1536,0.01221866657336553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,768,0.0035626664757728577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,1024,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,512,0.003185777821474605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,768,0.01204177737236023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,512,0.011171555353535546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,128,0.0028275555620590844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,256,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,64,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,128,0.010379555324713389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1536,32,0.0025253333151340485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,32,0.01037333326207267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,65536,0.025255999631351892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,65536,0.053767999013264976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,51200,0.02156266735659705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1536,64,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,51200,0.04359822140799629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,16384,0.011334222224023608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,12288,0.010758221977286868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,16384,0.021677333447668288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,10240,0.010006222460005019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,12288,0.018439999885029264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,8192,0.00908622228437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,10240,0.01678400072786543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,7168,0.0090515555606948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,8192,0.016006223029560514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,6144,0.008719111482302347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,7168,0.015226667126019796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,5120,0.007628444168302748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,6144,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,4096,0.007268444531493717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,4096,0.013764444324705334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,3584,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,3584,0.013767999907334646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,3072,0.006255110932721033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,5120,0.013788444300492605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,3072,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,2560,0.0052791109515561005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,2048,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,2560,0.01293155550956726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,1536,0.0053377776510185665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,2048,0.01182666669289271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,1024,0.004593777573770947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,1536,0.012225777738624148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,1024,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,768,0.0038942222793896994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,512,0.003642666671011183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,768,0.011780444118711682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,256,0.0032559999575217566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,512,0.011698666546079846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,128,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,256,0.010398222340477837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,64,0.0029288888391521242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,128,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,1024,32,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,64,0.010842666857772403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,65536,0.023540443844265405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,1024,32,0.01109688894616233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,51200,0.019648889700571697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,65536,0.0530426667796241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,16384,0.010520888699425591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,51200,0.04318044583002726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,16384,0.02051911089155409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,12288,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,10240,0.009718221922715506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,12288,0.017881777551439073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,8192,0.00831822223133511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,10240,0.016979555288950603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,7168,0.008719111482302347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,8192,0.01574844453069899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,6144,0.007990222010347579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,7168,0.015234667393896313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,6144,0.014226666755146451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,5120,0.007606222397751278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,4096,0.0069751110341813825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,5120,0.013779555757840475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,3584,0.006722666737106111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,4096,0.013863110707865821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,3072,0.006261333409282897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,3584,0.01310488912794325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,2560,0.005584888988071018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,3072,0.013099555340078143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,2560,0.012489777472284106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,2048,0.005155555489990446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,1536,0.005125333451562458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,1024,0.004376888689067629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,2048,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,768,0.0037342223028341928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,1536,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,1024,0.011186666786670685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,512,0.0036115555299652945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,768,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,256,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,512,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,256,0.010395555860466426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,128,0.010079111489984725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,64,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,64,0.010430222584141625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,768,32,0.0094355551732911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,65536,0.018629334039158292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,65536,0.051170667012532554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,51200,0.01667288939158122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,768,128,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,16384,0.009977777798970541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,51200,0.04256977637608846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,12288,0.009837333526876237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,16384,0.0203493336836497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,12288,0.017709333035680983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,10240,0.008989333278603023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,8192,0.008003555238246918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,10240,0.016590222716331482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,7168,0.00870222184393141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,8192,0.015936000479592215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,7168,0.015141333142916361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,6144,0.008026666939258575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,5120,0.007689778175618913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,6144,0.014467555615637036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,4096,0.006887110984987683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,5120,0.014340443743599786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,4096,0.013167111410035027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,3584,0.006653333289755716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,3072,0.006579555571079254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,3584,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,2560,0.0063031112982167145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,3072,0.01297511077589459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,2048,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,2560,0.012489777472284106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,1536,0.004522666749027041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,2048,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,1536,0.011887110769748688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,1024,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,768,0.003556444413132138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,1024,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,512,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,768,0.011902222202883827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,512,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,256,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,128,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,256,0.01036711119943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,64,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,128,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,512,32,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,64,0.010023111270533668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,65536,0.01661600006951226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,512,32,0.011038222246699862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,51200,0.015626667274369132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,65536,0.047910223404566445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,16384,0.008671999805503422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,16384,0.01998133295112186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,12288,0.007651555869314406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,12288,0.017855111095640395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,10240,0.0069848886794514125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,10240,0.0165955556763543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,8192,0.0063893331421746155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,8192,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,7168,0.006343110981914733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,7168,0.014868444866604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,6144,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,6144,0.014478221535682678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,5120,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,5120,0.013792888985739814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,4096,0.005929777605666055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,4096,0.013055110971132914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,3584,0.005555555638339784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,3584,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,51200,0.040001776483323835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,3072,0.0053688887920644544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,2560,0.005960888746711943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,2560,0.012425777812798819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,2048,0.005316444569163852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,3072,0.012919999659061432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,1536,0.004274666723277834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,2048,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,1024,0.003925333420435588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,1024,0.011034666664070554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,1536,0.0125937776433097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,768,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,512,0.0032879999942249725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,512,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,768,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,128,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,256,0.01036088913679123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,256,32,0.0029484445436133277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,64,0.010031110710567897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,128,0.011101333631409539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,65536,0.014130666851997375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,51200,0.01202755586968528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,256,32,0.010103999740547603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,16384,0.00794311116139094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,51200,0.03873066769705878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,12288,0.007143110864692264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,65536,0.04763377706209818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,16384,0.020260444945759244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,10240,0.007044444481531779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,12288,0.01755466726091173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,8192,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,10240,0.016899555921554565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,7168,0.00627377794848548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,8192,0.015504888362354703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,6144,0.00589155571328269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,7168,0.015138667490747241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,5120,0.005577777822812398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,6144,0.014480888843536377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,4096,0.0052133335007561584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,5120,0.01386666629049513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,3584,0.0059057776298787855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,4096,0.01313777764638265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,3072,0.0052915554907586836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,3584,0.013443555268976422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,2560,0.00602311102880372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,3072,0.012356444365448423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,2048,0.005208888815508948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,2560,0.012415111064910889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,1536,0.0042151109211974675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,2048,0.011339555184046427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,1024,0.003587555554178026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,1536,0.0120666664507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,768,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,1024,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,512,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,768,0.011400000088744693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,256,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,512,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,128,0.002793777734041214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,256,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,64,0.0025066667132907445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,128,0.010052444206343757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,128,32,0.0025031111306614345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,64,0.011026666396194033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,65536,0.010801778071456485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,128,32,0.009334222310119206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,51200,0.010321777727868822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,65536,0.046893331739637584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,16384,0.006952889263629913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,51200,0.038429333104027644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,12288,0.00683822234471639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,16384,0.020260444945759244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,10240,0.0069360000391801195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,12288,0.017535999417304993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,8192,0.005939555664857228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,10240,0.01682400041156345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,7168,0.006589333216349284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,8192,0.015423999892340766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,6144,0.005576000031497743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,7168,0.015208888385030957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,5120,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,6144,0.01447111037042406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,4096,0.005234666582610872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,5120,0.013754666679435305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,3584,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,4096,0.012852444416946836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,3072,0.0052373334765434265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,3584,0.013172444370057849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,2560,0.0059164443777667145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,3072,0.012114666402339935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,2048,0.005233777893914117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,2560,0.012402666939629449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,1536,0.004376000000370873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,2048,0.011343110766675739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,1024,0.0038204445607132385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,1536,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,768,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,1024,0.010865777730941772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,512,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,768,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,256,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,512,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,128,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,256,0.011227555572986603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,64,0.002509333400262727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,128,0.009973333113723332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,64,32,0.0026737778551048706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,64,0.011023999916182624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,64,32,0.009350222018029954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,65536,0.046296000480651855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,51200,0.009761778016885122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,51200,0.03839022252294753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,16384,0.006409777949253718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,16384,0.019880889190567862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,12288,0.006591111007663939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,12288,0.01754133403301239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,10240,0.006311111152172089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,65536,0.01076533314254549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,8192,0.006620444357395172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,10240,0.016532444291644625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,7168,0.0057653333577844836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,8192,0.0155395550860299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,6144,0.006266666783226862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,7168,0.01480977733929952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,5120,0.005225777626037598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,6144,0.014539556370841132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,4096,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,5120,0.013724444641007317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,3584,0.005602666487296422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,4096,0.013215111361609565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,3584,0.013357333011097379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,3072,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,3072,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,2560,0.005633777628342311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,2560,0.012425777812798819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,2048,0.005271111097600725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,2048,0.012445333103338877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,1536,0.004217777815130022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,1536,0.012103999654452005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,1024,0.0038417776425679526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,1024,0.011354666617181567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,768,0.0035031110876136353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,768,0.011384000380833944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,512,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,512,0.011360888679822287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,256,0.002809777855873108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,256,0.011225777367750803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,128,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,128,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,64,0.002500444443689452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,64,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,9,32,32,0.002504000026318762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,9,32,32,0.010040000081062317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,51200,0.9675333234998914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,51200,0.49337244033813477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,16384,0.3345422214931912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,16384,0.16682310899098715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,16384,0.18601777818467882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,12288,0.2480248875088162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,12288,0.14687111642625597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,10240,0.20918577247195771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,51200,0.5141350958082411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,10240,0.1608684460322062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,12288,0.14685600333743626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,8192,0.15164889229668513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,8192,0.11255999406178792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,7168,0.13380711608462864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,7168,0.10363822513156468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,10240,0.12681155734592015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,8192,0.10712622271643744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,6144,0.11528533034854466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,6144,0.09220533238516913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,7168,0.09639910856882732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,5120,0.09781333472993638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,5120,0.08049155606163873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,6144,0.08554666572146946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,4096,0.07771733072068956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,4096,0.06293866369459364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,5120,0.07663644684685601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,3584,0.06890755229526095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,3584,0.06259822183185153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,4096,0.06551733281877306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,3072,0.05970488654242622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,3072,0.049160887797673546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,2560,0.050535112619400024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,3584,0.060672885841793485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,3072,0.05657066901524862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,2560,0.04307555490069919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,2048,0.041356444358825684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,2560,0.0508266654279497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,2048,0.03690222236845229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,1536,0.03233155608177185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,1536,0.03230488962597317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,1024,0.0228426671690411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,2048,0.04751288890838623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,1536,0.04015200005637275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,768,0.018336888816621568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,1024,0.023204444183243647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,1024,0.03534311056137085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,768,0.022032888399230108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,512,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,768,0.031874666611353554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,256,0.008923555413881937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,512,0.03012711140844557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,256,0.01573777695496877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,128,0.006263111200597551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,512,0.017888888716697693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,256,0.026779555612140234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,64,0.005643555687533484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,128,0.013879111243618859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,65536,32,0.005571555760171678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,65536,128,0.020975111259354487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,64,0.01587466730011834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,65536,32,0.01520088811715444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,65536,0.5211333168877496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,65536,1.0834755367702908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,51200,0.8468444612291125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,51200,0.39942489729987246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,16384,0.2742648919423421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,16384,0.1366408930884467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,16384,0.15101422203911677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,12288,0.20126399728986952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,12288,0.12168533272213405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,10240,0.17070933183034262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,65536,0.5257226626078287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,10240,0.10556888580322266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,8192,0.14120266172620985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,12288,0.11960089206695557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,51200,0.4050622251298692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,8192,0.1050213310453627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,7168,0.12379466162787543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,10240,0.1047991116841634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,7168,0.09184266461266412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,6144,0.10906222131517197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,8192,0.08906666437784831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,6144,0.07947555515501234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,5120,0.08494933446248372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,7168,0.08107822471194796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,5120,0.07093599769804212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,6144,0.07265155845218234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,4096,0.07201688819461398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,5120,0.06586844391292997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,3584,0.060646222697363965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,4096,0.05518133441607157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,3584,0.05201511250601875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,4096,0.05563022030724419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,3072,0.052312887377209134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,3072,0.043135109874937266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,2560,0.045103109545177884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,3584,0.053242666853798755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,3072,0.050591998630099826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,2048,0.03738222188419766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,2560,0.042506668302747935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,2560,0.04626133375697666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,2048,0.031930665175120033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,1536,0.02940000096956889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,2048,0.04068888889418708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,1024,0.02130844526820713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,1536,0.02757599949836731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,1536,0.036087112294303045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,1024,0.020693333612547982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,768,0.0162462227874332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,1024,0.03494044476085239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,512,0.011014222270912595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,768,0.02035555574629042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,768,0.029112888707054987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,256,0.007254222200976477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,512,0.016943110360039603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,512,0.025916443930731878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,128,0.004575110971927643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,256,0.023733332753181458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,256,0.01348444405529234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,128,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,51200,128,0.02122933334774441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,64,0.003662222375472387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,51200,32,0.0035413333939181436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,64,0.013850666582584381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,51200,32,0.013990221752060784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,65536,0.3290986749860975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,65536,0.18187910980648467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,51200,0.25804710388183594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,51200,0.13871644602881536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,16384,0.09662933482064141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,16384,0.05425955520735847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,16384,0.06778399811850654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,12288,0.0694942209455702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,65536,0.1896604432000054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,12288,0.05964977873696221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,51200,0.15180444717407227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,10240,0.059676441881391734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,10240,0.05353422297371758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,12288,0.05678666962517632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,10240,0.05169333351982964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,8192,0.04695555567741394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,8192,0.04346399837070041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,7168,0.03681600093841553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,7168,0.03983644313282437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,7168,0.043325334787368774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,6144,0.03209244542651706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,6144,0.03590755661328634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,6144,0.040800889333089195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,8192,0.04984088738759359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,5120,0.027779555983013574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,4096,0.022727110319667395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,5120,0.03128711051411099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,5120,0.03907644417550828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,4096,0.025563556287023757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,3584,0.020414221617910597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,3584,0.025215110845035974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,3584,0.0332275562816196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,3072,0.018028444714016385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,3072,0.02200088898340861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,3072,0.030857778257793848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,2560,0.015455999308162265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,2560,0.021376889612939622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,2560,0.03016977839999729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,2048,0.013088888592190213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,2048,0.02771200074089898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,2048,0.018558222386572096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,1536,0.01037688884470198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,1536,0.01713244451416863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,1024,0.008164444731341468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,4096,0.034663110971450806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,1024,0.014110222458839417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,768,0.0069351109365622205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,1024,0.025528000460730657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,768,0.013780444032616086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,512,0.004782222211360931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,768,0.022251556317011516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,512,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,512,0.022016000416543748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,256,0.0032186667538351486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,1536,0.025703110628657874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,256,0.011735111474990845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,128,0.002895111011134254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,256,0.020301333732075162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,128,0.011362666885058085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,16384,128,0.018876444962289598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,64,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,16384,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,64,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,16384,32,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,65536,0.2438133292728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,65536,0.1546808878580729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,51200,0.19037332799699572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,51200,0.12034489048851861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,65536,0.15344266096750894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,16384,0.07342844539218478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,51200,0.12207910749647354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,16384,0.04814844330151876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,12288,0.051290667719311185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,12288,0.051627556482950844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,16384,0.05877333217196994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,10240,0.04447466797298855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,12288,0.049954666031731494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,10240,0.04637066523234049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,8192,0.037271999650531344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,10240,0.04564177658822802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,8192,0.03795288999875387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,7168,0.03285155693689982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,8192,0.04154577851295471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,7168,0.03475644522243076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,6144,0.028466665082507666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,7168,0.038677334785461426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,6144,0.0306000014146169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,5120,0.025495999389224585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,6144,0.03700266612900628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,5120,0.027181333965725366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,4096,0.020822222034136455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,5120,0.03394133183691237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,4096,0.03156000044610765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,3584,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,3584,0.02299733294381036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,3584,0.030127998855378892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,3072,0.01699911057949066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,3072,0.020308444897333782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,3072,0.029342220889197454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,2560,0.015119999647140503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,4096,0.02440888848569658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,2560,0.019912888606389362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,2048,0.010624888870451184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,2560,0.028901331954532202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,2048,0.01759022143152025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,2048,0.027570666538344488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,1536,0.008722667064931657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,1536,0.015799111790127225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,1536,0.025474665893448725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,1024,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,768,0.006305777778228124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,1024,0.013449778159459433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,1024,0.023139556248982746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,512,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,768,0.02200088898340861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,512,0.012100444071822695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,256,0.0032631111227803757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,512,0.02198933396074507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,256,0.010821333361996545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,256,0.02200800014866723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,128,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,12288,128,0.01991555591424306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,64,0.0034799998005231223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,64,0.012119111087587146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,32,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,32,0.011143111520343356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,12288,128,0.002971555623743269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,65536,0.20377866427103677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,65536,0.1486275593439738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,51200,0.16297954983181423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,12288,768,0.013732444081041547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,51200,0.11571378178066677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,16384,0.06292355722851224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,16384,0.046450667911105685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,65536,0.13542400466071233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,12288,0.04387199878692627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,12288,0.043021334542168506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,16384,0.054478221469455294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,10240,0.037999110089408025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,51200,0.11003466447194417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,12288,0.047404444879955716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,8192,0.031204445494545832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,10240,0.044623110029432506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,8192,0.033369776275422834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,7168,0.02864711152182685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,7168,0.029464887248145208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,7168,0.03718400001525879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,6144,0.025645333859655593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,6144,0.026548445224761963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,10240,0.037419554260041975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,5120,0.022083555658658344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,6144,0.03560533457332187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,8192,0.040022220876481794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,5120,0.024069334069887798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,4096,0.018406222263971966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,5120,0.03353688783115811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,4096,0.02163822286658817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,3584,0.017237333787812125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,4096,0.03199911117553711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,3584,0.02068088948726654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,3584,0.030979557169808283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,3072,0.019495111372735765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,3072,0.030631999174753826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,2560,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,2560,0.018617777360810172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,3072,0.015816888875431485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,2048,0.010093332992659675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,2560,0.029579556650585596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,2048,0.016552888684802584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,2048,0.028361777464548748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,1536,0.008547555241319869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,1536,0.025916443930731878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,1536,0.015512888630231222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,1024,0.006978666616810693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,1024,0.022515555222829182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,1024,0.013431111143694984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,768,0.0053048887186580235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,768,0.013219555219014486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,512,0.004331555631425646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,512,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,512,0.022413333257039387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,256,0.0031635556370019913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,256,0.011869333684444427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,768,0.022322666313913133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,128,0.0028515555378463534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,256,0.022056889202859666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,128,0.01112711098459032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,10240,128,0.021037333541446265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,64,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,10240,32,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,32,0.010817777779367236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,10240,64,0.012083555261294046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,65536,0.16480799516042074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,65536,0.09930844439400567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,51200,0.13140532705518934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,51200,0.07803199688593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,65536,0.12307911449008518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,16384,0.049738665421803795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,16384,0.034930666287740074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,51200,0.10091822014914618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,12288,0.032764444748560585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,16384,0.05106222298410204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,12288,0.045547554890314736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,12288,0.03465422325664096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,10240,0.028456889920764502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,10240,0.031924443112479314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,10240,0.041730665498309664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,8192,0.023376888699001733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,8192,0.02718044486310747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,7168,0.02071466710832384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,8192,0.03823733329772949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,7168,0.025072889195548162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,7168,0.036905778778923884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,6144,0.018328888548745047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,6144,0.034685333569844566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,6144,0.02316088808907403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,5120,0.016201777590645682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,5120,0.03299288948376974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,4096,0.012921777864297232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,4096,0.01887822151184082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,4096,0.030750221676296655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,3584,0.011686222420798408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,3584,0.017966222431924608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,3584,0.030232889784706965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,3072,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,5120,0.02064622276359134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,3072,0.016899555921554565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,2560,0.009461333354314169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,3072,0.029873778422673542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,2560,0.016195555528004963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,2560,0.02920444475279914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,2048,0.00851822230550978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,1536,0.007148444652557373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,2048,0.014840889308187695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,2048,0.027845333019892376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,1536,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,1024,0.005081777771313985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,1024,0.012004444168673621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,768,0.003839111162556542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,1024,0.023066666391160753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,768,0.012124444047609964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,768,0.02270577847957611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,512,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,512,0.012136889000733694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,512,0.020784889658292133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,256,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,256,0.022335110439194575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,256,0.010769777827792697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,1536,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,128,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,64,0.002699555622206794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,8192,128,0.01959111127588484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,128,0.010445333189434476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,8192,32,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,64,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,8192,32,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,65536,0.14460799429151747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,65536,0.09057422478993733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,51200,0.11609689394632976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,51200,0.07114222314622667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,65536,0.12042310502794053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,16384,0.04240977764129639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,16384,0.0326551099618276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,16384,0.05058755477269491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,12288,0.03147111005253262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,51200,0.10023378001319037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,12288,0.033287998702791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,10240,0.027190221680535212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,12288,0.04389244318008423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,10240,0.030032889710532293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,8192,0.023149333066410486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,10240,0.041272888580958046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,8192,0.026025777061780293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,8192,0.03696977761056688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,7168,0.024029332730505202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,7168,0.035657776726616755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,6144,0.016885333591037326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,6144,0.021573333276642695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,6144,0.03360711203681098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,5120,0.0148062225845125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,7168,0.020766221814685397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,5120,0.020240000552601285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,5120,0.03271111183696323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,4096,0.012190221912331052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,4096,0.01861422260602315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,4096,0.03050755461057027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,3584,0.011206222077210745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,3584,0.0299271113342709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,3584,0.017651556266678702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,3072,0.009964444571071202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,3072,0.028880887561374243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,3072,0.01647911138004727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,2560,0.009301332963837517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,2560,0.015268445014953613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,2560,0.027561777167850073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,2048,0.008277333445019191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,2048,0.014325333966149224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,2048,0.02604088849491543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,1536,0.006888888776302338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,1536,0.022993778189023335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,1536,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,1024,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,1024,0.011811555259757571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,1024,0.022389334109094407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,768,0.0038764443662431505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,768,0.012011555333932241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,768,0.021330666210916307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,512,0.0032479998966058097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,512,0.021701332595613267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,512,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,256,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,256,0.010696888797812991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,256,0.02141511109140184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,128,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,128,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,7168,128,0.01958488921324412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,64,0.002540444541308615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,7168,32,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,64,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,65536,0.12616088655259874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,7168,32,0.010701333483060202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,51200,0.1017200019624498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,65536,0.08164355489942762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,51200,0.06485778093338013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,16384,0.040157334672080144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,16384,0.030555556217829388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,65536,0.11980177296532525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,51200,0.09889422522650824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,12288,0.029280887709723577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,16384,0.050373332368003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,12288,0.030086222622129653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,12288,0.043225778473748096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,10240,0.024703111913469102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,8192,0.02065333392884996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,10240,0.028555555476082697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,10240,0.03912711143493652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,7168,0.018750222192870248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,8192,0.02437422176202138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,8192,0.03666844301753574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,7168,0.023345778385798138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,7168,0.03533066643608941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,6144,0.017141333884663053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,5120,0.014843554960356818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,6144,0.034296890099843345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,5120,0.019665777683258057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,4096,0.011271110839313932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,5120,0.032279110617107816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,4096,0.030206223328908283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,4096,0.01791911158296797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,3584,0.010031110710567897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,3584,0.01717066599263085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,3584,0.02985422147644891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,3072,0.009561777942710454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,3072,0.01592177814907498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,6144,0.021551999780866835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,3072,0.02850133180618286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,2560,0.008647110727098253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,2560,0.01516711049609714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,2560,0.026750221848487854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,2048,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,2048,0.01365600029627482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,2048,0.024246222443050806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,1536,0.006574222197135289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,1536,0.012744888663291931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,1536,0.021662222014533147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,1024,0.004311999926964442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,1024,0.022320000661744013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,1024,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,768,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,768,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,768,0.021634666456116572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,512,0.0032471112079090546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,512,0.020118221640586853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,512,0.011420444481902652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,256,0.0031911111954185697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,256,0.011115555961926779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,256,0.01955911186006334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,128,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,128,0.010037333601050908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,6144,128,0.019294222195943195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,64,0.0025724443710512584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,6144,32,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,64,0.011221333510345884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,6144,32,0.00998755544424057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,65536,0.10572089089287652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,65536,0.0735644433233473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,51200,0.08407644430796306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,51200,0.05919288926654392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,65536,0.1193315585454305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,16384,0.03503733211093479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,16384,0.028389334678649902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,51200,0.09878577788670857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,12288,0.0255422227912479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,16384,0.0495324432849884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,12288,0.030504888958401147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,10240,0.021754667162895203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,12288,0.04342755675315857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,10240,0.025786666406525507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,10240,0.040038221412234835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,8192,0.018975999620225694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,8192,0.022720888257026672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,7168,0.01701955497264862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,7168,0.021546666820844013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,7168,0.03524444500605265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,6144,0.015167999598715039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,6144,0.020262221495310467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,6144,0.03406755460633172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,5120,0.013103110922707452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,8192,0.03635822070969476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,5120,0.018925334016482036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,4096,0.010016000105275048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,5120,0.03138044476509094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,4096,0.017521777086787753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,4096,0.029496888319651287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,3584,0.009907555249002244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,3072,0.009168888959619734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,3584,0.01647733317481147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,3584,0.028371555937661067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,2560,0.007630222373538547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,3072,0.02851111027929518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,2560,0.014217777384652032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,2048,0.007123555574152205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,2560,0.026829333768950567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,2048,0.013311111264758639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,2048,0.024012444747818842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,1536,0.0057591112951437635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,1536,0.021651556094487507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,1024,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,1024,0.011703111231327057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,1024,0.021008888880411785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,768,0.00369422220521503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,3072,0.01514844430817498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,768,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,512,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,768,0.022685334086418152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,512,0.011332444846630096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,256,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,1536,0.012461333639091916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,512,0.022469333476490442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,256,0.01033688916100396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,128,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,256,0.021006221572558086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,128,0.010053333308961656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,5120,32,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,64,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,5120,32,0.009584888815879822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,65536,0.08535555336210464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,65536,0.06724622514512804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,51200,0.0682142244444953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,51200,0.05474844574928284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,65536,0.11924089325798883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,5120,128,0.020308444897333782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,51200,0.09885866774453057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,16384,0.027979556057188246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,16384,0.025213332639800176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,12288,0.020996444755130347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,16384,0.04876266585456001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,12288,0.025384889708624944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,10240,0.018237334158685472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,10240,0.02493155499299367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,12288,0.042565332518683545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,8192,0.015187554889255099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,10240,0.03985688752598233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,8192,0.02314844396379259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,7168,0.014157333307796054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,8192,0.03710755705833435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,7168,0.034973333279291786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,7168,0.020573332905769348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,6144,0.012875555290116204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,6144,0.019467555814319186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,6144,0.033320890532599554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,5120,0.01186044431395001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,5120,0.017909333109855652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,4096,0.01053244454993142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,5120,0.032186667124430336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,4096,0.016536888149049547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,4096,0.030672887961069744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,3584,0.011495999991893768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,3584,0.028992888000276353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,3072,0.009457777771684859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,3072,0.01386755539311303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,3072,0.027863999207814533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,2560,0.00869066682126787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,2560,0.013437333206335703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,2560,0.02492888934082455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,2048,0.007333333293596904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,3584,0.016376000311639573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,2048,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,1536,0.005993777679072485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,2048,0.024454222785101995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,1536,0.012310222619109683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,1536,0.023378666904237535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,1024,0.004916444420814514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,1024,0.02231377859910329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,768,0.004538666870858935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,768,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,768,0.02160888910293579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,512,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,512,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,512,0.021372444099850122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,256,0.0031724443866146933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,1024,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,256,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,128,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,256,0.021915555000305176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,128,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,64,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,4096,128,0.020263110597928364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,4096,32,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,64,0.0110871113008923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,65536,0.07541866434945001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,4096,32,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,65536,0.06420711014005873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,51200,0.06079644627041287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,65536,0.11894222100575765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,51200,0.05237866772545708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,16384,0.024078221784697637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,16384,0.02435644467671712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,51200,0.09904622369342381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,12288,0.01848266687658098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,16384,0.048473778698179454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,12288,0.025263999899228413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,10240,0.01629688839117686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,12288,0.043037331766552396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,10240,0.024559999505678814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,10240,0.03945777813593546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,8192,0.013949332965744866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,8192,0.02031822171476152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,8192,0.036679112248950534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,7168,0.013548444542619916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,6144,0.012514666550689273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,7168,0.03536266750759549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,6144,0.03389511174625821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,6144,0.01827111177974277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,5120,0.01147466649611791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,7168,0.020649777518378366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,4096,0.010044444766309526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,5120,0.017229333519935608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,5120,0.03194311261177063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,3584,0.010381333529949188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,4096,0.02959022257063124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,4096,0.015504888362354703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,3072,0.008966222405433655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,3584,0.02882400155067444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,3584,0.015607110328144498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,2560,0.008624888956546783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,3072,0.01406133340464698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,3072,0.026075555218590632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,2048,0.0069422221018208405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,2560,0.014182221558358936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,2048,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,2048,0.023390221926901076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,1536,0.00600000015563435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,1536,0.012305777933862476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,1536,0.022688888841205176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,1024,0.004892444445027245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,1024,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,2560,0.02471911079353756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,768,0.0042071110672420925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,1024,0.02197244432237413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,768,0.021712000171343487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,768,0.01146133326821857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,512,0.005277333160241445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,512,0.021384888225131567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,512,0.0107351111041175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,256,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,256,0.021347555849287245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,256,0.01038933296998342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,128,0.0028239999794297745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3584,128,0.019260444574885897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,128,0.010451555252075195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,64,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3584,32,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,64,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3584,32,0.010105777945783403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,65536,0.06536266538831922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,65536,0.06183111005359226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,51200,0.050534221861097545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,65536,0.11871466371748184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,51200,0.050395554966396756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,16384,0.021213332811991375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,16384,0.023670223024156358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,51200,0.09827288654115464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,12288,0.01644444465637207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,16384,0.0485226677523719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,12288,0.04305511050754123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,10240,0.014682667122946845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,10240,0.02233155568440755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,10240,0.03916800022125244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,8192,0.012877333495352002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,8192,0.019289778338538278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,8192,0.03689333465364244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,12288,0.023725334140989516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,7168,0.019088000059127808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,7168,0.034627556800842285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,6144,0.010841777755154504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,6144,0.017921777235137094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,6144,0.03267200125588311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,7168,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,5120,0.009934222532643212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,5120,0.01680799987581041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,4096,0.009343999955389235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,5120,0.03016177813212077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,4096,0.015378667248619927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,4096,0.02853511108292474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,3584,0.009018667042255402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,3584,0.0281315545241038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,3584,0.014142221874660917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,3072,0.008224888808197444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,3072,0.025496888491842482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,3072,0.013456000222100152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,2560,0.007328888608349695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,2560,0.024715556038750544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,2560,0.013050666285885705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,2048,0.005952888892756567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,2048,0.023688889212078516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,2048,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,1536,0.005959110955397288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,1536,0.022336888644430373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,1536,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,1024,0.004929777648713854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,1024,0.02203911046187083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,1024,0.011561777856614856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,768,0.004302222281694412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,768,0.02167111138502757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,768,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,512,0.003931555483076307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,512,0.021690666675567627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,512,0.010741333166758219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,256,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,256,0.021055110626750525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,256,0.010400888820489248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,128,0.0029279999434947968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,64,0.0026933333526055017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,128,0.010236444572607676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,3072,128,0.019296889503796894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,3072,32,0.002570666579736604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,65536,0.05559911330540975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,32,0.010401777923107147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,65536,0.05937777625189888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,51200,0.04580266608132256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,65536,0.11848978201548259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,51200,0.04877866639031304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,16384,0.020075556304719713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,16384,0.023071999351183575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,51200,0.0986515548494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,3072,64,0.011306666665607028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,12288,0.01463822192615933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,16384,0.04689866635534498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,10240,0.013839999834696451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,12288,0.02197244432237413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,12288,0.04184444414244758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,10240,0.021700445148679946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,8192,0.011971555650234222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,10240,0.039157334301206805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,8192,0.018594667315483093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,8192,0.03601866629388597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,7168,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,7168,0.0350213348865509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,6144,0.010614222122563256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,6144,0.016587555408477783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,6144,0.03330577744377984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,5120,0.00944622192117903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,5120,0.015637333194414776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,5120,0.03163466519779629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,7168,0.017986666825082567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,4096,0.014119111829333834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,4096,0.02721600068940057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,3584,0.00831733312871721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,3584,0.013618666264745923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,4096,0.008877333667543199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,3072,0.008096000386608971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,3584,0.02717688845263587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,3072,0.013167999684810638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,2560,0.007104000283612146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,3072,0.025426665941874187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,2560,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,2560,0.024417777856190998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,2048,0.006190222170617845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,2048,0.023743111226293776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,2048,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,1536,0.005603555589914322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,1536,0.022696000006463792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,1536,0.011976888610257043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,1024,0.004638222356637319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,1024,0.02203377750184801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,1024,0.011406222151385413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,768,0.00423911131090588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,768,0.021380444367726643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,768,0.011416888899273343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,512,0.003899555653333664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,512,0.021606221795082092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,512,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,256,0.011156444748242697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,256,0.02100533412562476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,128,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2560,128,0.019263111882739596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,128,0.01037333326207267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,64,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2560,32,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,64,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2560,32,0.010142222046852112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,65536,0.045532445112864174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,65536,0.057384888331095375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,51200,0.03779822256830003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,65536,0.11826488706800674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,51200,0.04731111062897576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,16384,0.015000889698664347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,51200,0.09837866491741604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,16384,0.02277333372169071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,12288,0.012838222086429596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,16384,0.04708800050947401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,12288,0.0206675562593672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,10240,0.01164533363448249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,12288,0.04153955645031399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,10240,0.019688889384269714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,8192,0.011219555305110084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,10240,0.03920266694492764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,8192,0.017626666360431247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,7168,0.01036533299419615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,8192,0.035791112316979304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,7168,0.016207110550668504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,6144,0.009680888719028896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,6144,0.014871999621391296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,6144,0.032961779170566134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,5120,0.009092444347010719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,5120,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,5120,0.029559112257427637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,4096,0.008014221986134848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,7168,0.03264000018437704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,4096,0.01384800010257297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,3584,0.007312888900438945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,4096,0.027286221583684284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,3584,0.013365333278973898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,3072,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,3584,0.02645066711637709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,3072,0.012830221818553077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,3072,0.025420443879233465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,2560,0.005933333188295364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,2560,0.024400000770886738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,2560,0.012458666331238218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,2048,0.006630222416586346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,2048,0.023825777901543513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,2048,0.011757333245542316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,1536,0.005669333454635408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,1536,0.022410665949185688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,1536,0.011853333148691388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,1024,0.004951999833186467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,1024,0.022408889399634466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,1024,0.011503111157152386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,768,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,768,0.01128266668981976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,768,0.021681777305073206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,512,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,512,0.01072622256146537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,512,0.020997333857748244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,256,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,256,0.010388444695207808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,256,0.021027555068333943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,128,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,128,0.01036888857682546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,2048,128,0.019296000401178997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,64,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,64,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,2048,32,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,2048,32,0.010065777434243096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,65536,0.035832888550228544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,65536,0.05570133527119955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,65536,0.11847466892666286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,51200,0.02879200047916836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,51200,0.0459591117170122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,51200,0.09795200162463719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,16384,0.01313333296113544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,16384,0.022143110632896423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,16384,0.04699110984802246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,12288,0.011233777635627322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,12288,0.02032444377740224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,12288,0.04087022278043959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,10240,0.01404888927936554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,10240,0.019144889381196763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,10240,0.037999110089408025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,8192,0.012480888929631976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,8192,0.016896888613700867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,8192,0.0346995559003618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,7168,0.011723555624485016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,7168,0.015899555550681222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,7168,0.033257779147889875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,6144,0.008427555362383524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,6144,0.014836443795098199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,6144,0.03087200058831109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,5120,0.007654222349325816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,5120,0.028516444895002577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,4096,0.007650666766696506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,4096,0.01352355546421475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,4096,0.02720977697107527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,3584,0.007000889215204451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,3584,0.013179555535316467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,5120,0.014035556051466199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,3584,0.026173333326975506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,3072,0.012520000338554382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,3072,0.02551288902759552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,2560,0.005620444400442972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,2560,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,2560,0.024166221419970196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,3072,0.006380444599522485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,2048,0.005934222290913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,2048,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,2048,0.023445333043734234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,1536,0.005006222261322869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,1536,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,1536,0.021376889612939622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,1024,0.004287111262480418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,1024,0.011501332951916588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,1024,0.021356445219781663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,768,0.003937777959638172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,768,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,768,0.020392888122134738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,512,0.0032151111712058387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,512,0.011122666299343109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,512,0.020986666282018025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,256,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,256,0.019626667102177937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,256,0.01036533299419615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,128,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1536,128,0.017920888132519193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,128,0.010392000277837118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1536,32,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,64,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1536,32,0.009422221945391761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,65536,0.025215110845035974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,65536,0.05369777811898125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,51200,0.020922667450375024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,65536,0.11815288331773545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,51200,0.042987555265426636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,16384,0.013520888984203339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,51200,0.09751200013690525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,16384,0.021228444245126512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,12288,0.011720000041855706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,16384,0.045999109745025635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,12288,0.017837334010336135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,12288,0.04032266802257962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,10240,0.01069777790043089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,10240,0.03693066702948676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,10240,0.016865778300497267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,8192,0.009717333647939894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,8192,0.03429066803720262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,8192,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,7168,0.009359111388524374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,7168,0.015120888749758402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,7168,0.03220622075928582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,6144,0.007659555309348636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,6144,0.014524444937705994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,5120,0.007028444773621029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,5120,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,6144,0.031341334184010826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,4096,0.0069439998931354946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,5120,0.02849510974354214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,4096,0.01349422252840466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,3584,0.00685511115524504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,4096,0.02681244413057963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,3584,0.013129777378506131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,3584,0.026145777768558923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,3072,0.006311111152172089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,3072,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,3072,0.026477333572175767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,2560,0.005631999837027655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,2560,0.02414488957987891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,2560,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,2048,0.006311999840868845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,2048,0.023215999205907185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,1536,0.005683555371231503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,2048,0.012314666476514606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,1536,0.011734222372372946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,1024,0.004576888763242298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,1536,0.023567110300064087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,1024,0.011391110718250275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,768,0.0038515557017591265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,1024,0.02288088864750332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,768,0.01110400011142095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,512,0.0035182221068276297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,768,0.022755554980701868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,512,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,256,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,512,0.022073777185546026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,128,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,256,0.011350221931934357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,256,0.02072177827358246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,64,0.0028231110837724474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,128,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,1024,128,0.019362666540675692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,1024,32,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,64,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,65536,0.021551110678248938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,1024,32,0.010097777677906884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,51200,0.01752622259987725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,65536,0.052453332477145724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,65536,0.11847110589345296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,51200,0.0429111123085022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,51200,0.09747288624445598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,16384,0.012522666818565793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,12288,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,16384,0.02030311193731096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,16384,0.04596177736918131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,10240,0.009167110754383935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,12288,0.04093599981731839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,12288,0.018279999494552612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,10240,0.016891555653678048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,8192,0.009136888715955947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,10240,0.03673422336578369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,7168,0.008375999828179678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,8192,0.01555377741654714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,8192,0.03446133269204034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,7168,0.01520088811715444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,6144,0.007332444190979004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,7168,0.03224888775083754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,5120,0.006590222318967183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,6144,0.014848000473446317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,6144,0.031025777260462444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,5120,0.029286222325430974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,5120,0.014173333843549093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,4096,0.007339555356237624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,3584,0.0070186663005087115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,4096,0.013526221944226159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,4096,0.027106665902667578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,3072,0.006605333338181178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,3584,0.026499556170569524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,3584,0.013168888787428537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,3072,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,2560,0.006322666588756774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,3072,0.02581600017017788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,2048,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,2560,0.012177777787049612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,2560,0.025928000609079998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,2048,0.012120889292822944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,1536,0.005046222358942032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,2048,0.023023110296991136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,1024,0.004228444563017951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,1536,0.01178311142656538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,1536,0.023464888334274292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,1024,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,1024,0.022038223014937505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,768,0.021746666895018682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,768,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,512,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,512,0.01037333326207267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,256,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,256,0.010404444403118556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,256,0.020604444874657523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,768,0.003839111162556542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,128,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,128,0.011041777829329172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,512,0.02139644490347968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,64,0.002521777732504739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,768,128,0.02033155494266086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,768,32,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,64,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,65536,0.02200177808602651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,768,32,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,65536,0.1186435619990031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,51200,0.018983110785484314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,51200,0.04305244485537211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,65536,0.05138577686415779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,16384,0.010374222364690568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,16384,0.020285333196322124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,16384,0.046057777272330395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,12288,0.008431110945012834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,12288,0.017997332745128207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,12288,0.04044888748062982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,10240,0.007721777591440413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,51200,0.09784977965884739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,10240,0.016917333006858826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,10240,0.037752889924579196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,8192,0.008002666963471307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,8192,0.034303112162484065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,8192,0.01595288846227858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,7168,0.007684444387753804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,7168,0.032297776805029974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,6144,0.0069884442620807225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,7168,0.015597333510716757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,6144,0.014803555276658801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,5120,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,6144,0.03130044539769491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,5120,0.014159111513031853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,4096,0.00700711127784517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,5120,0.029496888319651287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,4096,0.01347555551264021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,3584,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,3584,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,4096,0.028314666615592107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,3072,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,3584,0.02645333276854621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,3072,0.025753777888086107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,3072,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,2560,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,2560,0.025748444928063288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,2048,0.005858666780922148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,2048,0.012333333492279053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,2560,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,1536,0.004295111116435793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,2048,0.024417777856190998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,1536,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,1536,0.0236791107389662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,1024,0.003866666720973121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,1024,0.022118222382333543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,768,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,768,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,768,0.021673777037196692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,512,0.0031759999692440033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,512,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,512,0.021359110871950786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,1024,0.011887110769748688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,256,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,128,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,256,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,128,0.011048888994587792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,64,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,512,128,0.02032355633046892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,512,32,0.002961777771512667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,64,0.011152889165613385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,65536,0.014871999621391296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,65536,0.0494044456217024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,512,32,0.010859555668301053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,51200,0.013354666531085968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,65536,0.11795111497243245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,51200,0.03992444607946608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,16384,0.00868622213602066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,51200,0.09627288579940796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,12288,0.008343111309740279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,16384,0.045257776975631714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,16384,0.020199111766285367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,10240,0.007350222104125553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,12288,0.039805332819620766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,10240,0.01662400033738878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,8192,0.0067528887755341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,10240,0.037674665451049805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,8192,0.015496888094478183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,8192,0.03429333368937174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,7168,0.00656088896923595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,7168,0.03124089042345683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,6144,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,7168,0.01556799974706438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,6144,0.014434667097197639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,5120,0.005995555470387141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,12288,0.017904000149832833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,6144,0.030858665704727173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,4096,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,5120,0.029530667596393164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,5120,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,4096,0.01348088930050532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,3584,0.006263999889294307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,4096,0.02641955514748891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,3584,0.025058666865030926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,3584,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,3072,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,3072,0.024494222468800012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,2560,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,2560,0.011834666960769229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,2560,0.023544889357354905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,2048,0.005693333430422678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,2048,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,2048,0.02270488937695821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,1536,0.004298666699065102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,3072,0.01316088851955202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,1536,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,1024,0.003677333394686381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,1536,0.023958222733603582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,1024,0.011176889141400656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,768,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,1024,0.022631110416518316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,768,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,512,0.003167111012670729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,768,0.02287111183007558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,512,0.01074488874938753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,256,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,512,0.02223555578125848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,256,0.010119111173682744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,128,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,256,0.02165688905451033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,128,0.010096888575288985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,64,0.002532444480392668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,256,128,0.020536889632542927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,64,0.010270222193664974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,256,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,65536,0.013464888764752282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,256,32,0.009367999931176504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,65536,0.04745511213938395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,65536,0.11561866601308186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,51200,0.012443555725945367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,51200,0.038443555434544883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,16384,0.007663999994595845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,51200,0.09621422158347236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,16384,0.01976444489426083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,12288,0.007106666763623555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,16384,0.04560000035497877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,12288,0.017587555779351127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,10240,0.006671111202902264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,12288,0.039814223845799766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,10240,0.016564443707466125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,8192,0.006326222171386083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,10240,0.03773155477311876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,8192,0.015473778049151102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,8192,0.03428355521625943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,7168,0.015486222174432544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,7168,0.032827556133270264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,6144,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,6144,0.01462933255566491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,7168,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,5120,0.00554044461912579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,5120,0.013826666606797112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,5120,0.028551999065611098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,4096,0.005408000200986862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,4096,0.012844444149070315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,4096,0.026206221845414903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,3584,0.005704000178310607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,6144,0.031223111682467993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,3584,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,3072,0.00563822231358952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,3584,0.02720622221628825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,2560,0.006105777704053455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,3072,0.012149333126015134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,3072,0.02644355595111847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,2560,0.02370400064521366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,2048,0.006011555592219035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,2560,0.012076444096035428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,1536,0.004212444441186057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,2048,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,2048,0.024466666910383437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,1536,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,1536,0.020974222156736586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,1024,0.003768889026509391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,768,0.003552000141806073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,1024,0.020909332566791113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,1024,0.011056889262464313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,512,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,768,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,768,0.019966221517986722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,256,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,512,0.010597333312034607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,512,0.02202933364444309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,128,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,256,0.010105777945783403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,256,0.020625778370433383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,64,0.0025128889828920364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,8,128,128,0.01622577839427524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,128,0.011019555230935415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,128,32,0.002549333290921317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,65536,0.010455999937322406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,64,0.010737777584128909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,128,32,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,51200,0.009376000199053023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,65536,0.04660177893108792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,16384,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,51200,0.03933599922392104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,12288,0.0063484443558586975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,16384,0.019976889093716938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,10240,0.005562666803598404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,12288,0.018704889549149405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,10240,0.016619554824299283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,8192,0.015473778049151102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,7168,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,7168,0.015102222561836243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,6144,0.005962666538026597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,6144,0.014278221461508008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,5120,0.005959110955397288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,8192,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,4096,0.005986666513813867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,5120,0.013919111755159167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,3584,0.005791999813583162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,4096,0.01350044459104538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,3584,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,3072,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,2560,0.00598933340774642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,3072,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,2048,0.005623111294375525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,2560,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,1536,0.004243555582231945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,2048,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,1024,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,1536,0.012156444291273752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,768,0.003576000117593341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,1024,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,512,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,768,0.011733333269755045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,256,0.002942222274012036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,512,0.01072266697883606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,128,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,256,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,64,0.0026764445420768526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,128,0.010059555371602377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,64,0.01148888882663515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,64,32,0.0025395556456512874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,65536,0.008991111483838823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,64,32,0.010814222196737925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,51200,0.008513777620262569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,51200,0.039128889640172325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,16384,0.006632888896597757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,16384,0.020273778173658583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,12288,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,12288,0.017921777235137094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,10240,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,65536,0.046592000457975596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,8192,0.005511111269394557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,10240,0.016872000363137987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,7168,0.006329777754015393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,8192,0.015493333339691162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,6144,0.005621333503060871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,7168,0.015404444601800708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,5120,0.005917333480384614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,6144,0.014187556174066333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,4096,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,5120,0.014138667119873894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,3584,0.005648889061477449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,4096,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,3072,0.005254222287072076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,3584,0.013175110850069257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,2560,0.005927111125654644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,3072,0.01238933371173011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,2048,0.0053075556125905775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,2560,0.012477333347002665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,1536,0.004516444272465176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,2048,0.011695110963450538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,1024,0.003612444632583194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,1536,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,768,0.0034835553831524322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,1024,0.01073066641887029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,512,0.003157333367400699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,768,0.01143555591503779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,512,0.011060444845093621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,256,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,128,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,256,0.01038755559258991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,64,0.002522666628162066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,128,0.011037333144081963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,8,32,32,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,64,0.010684444672531553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,8,32,32,0.010748444332016839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,51200,0.492961777581109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,51200,0.9663955900404188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,16384,0.33497601085238987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,12288,0.24645421240064833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,16384,0.24254488945007324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,12288,0.1792000002331204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,10240,0.15743466218312582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,10240,0.20822132958306205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,8192,0.1699919965532091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,7168,0.13352621926201716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,8192,0.11911022663116455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,7168,0.0988159974416097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,6144,0.11526400513119167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,5120,0.09780266549852158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,6144,0.09177866909239028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,4096,0.07775821950700548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,5120,0.07997955878575642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,3584,0.06870844629075792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,4096,0.06809777683681913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,3072,0.05973777506086561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,3584,0.057836446497175426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,2560,0.05039466751946343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,3072,0.052449779378043286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,2048,0.04142933421664768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,2560,0.044844445255067616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,1536,0.032186667124430336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,2048,0.041858666472964816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,1024,0.022775999373859827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,1536,0.03387377659479777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,768,0.018532445033391316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,1024,0.025753777888086107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,512,0.013247110777431063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,768,0.022865777214368183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,256,0.008767111433876885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,512,0.018812444474962022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,128,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,256,0.015152888165579902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,64,0.005575110928879843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,128,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,65536,32,0.0052986666560173035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,64,0.015191111299726697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,65536,32,0.014478221535682678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,65536,0.5201288859049479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,65536,1.0789751476711698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,51200,0.8473715782165527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,51200,0.39888978004455566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,16384,0.27294667561848956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,12288,0.2007733318540785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,16384,0.1863457759221395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,10240,0.17014755143059623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,10240,0.12072711520724827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,12288,0.1410026682747735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,8192,0.13980889320373535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,8192,0.09839022159576416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,7168,0.12299644947052002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,6144,0.10854844252268474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,7168,0.08124088578754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,5120,0.09250044160419041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,6144,0.07624089055591159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,4096,0.07140355639987521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,5120,0.06764266888300578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,3584,0.06086577971776327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,3584,0.04875822199715508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,4096,0.058202664057413735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,3072,0.05244622296757168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,2560,0.04429866539107429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,3072,0.0439280006620619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,2048,0.03728444377581278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,2560,0.03896177808443705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,1536,0.028808888461854722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,2048,0.03664977682961358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,1024,0.021292444732454088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,1536,0.028753777345021565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,768,0.01608000033431583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,1024,0.022720000810093347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,512,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,768,0.0206631107462777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,256,0.007342221836249034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,512,0.01715377800994449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,128,0.0038364442686239877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,256,0.014495111174053617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,64,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,128,0.013096888860066732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,51200,32,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,64,0.014200000299347771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,51200,32,0.013601777454217276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,65536,0.3288497659895155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,65536,0.18119643794165718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,51200,0.25823378562927246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,16384,0.09510666794247097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,51200,0.13872978422376844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,12288,0.06888710790210299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,16384,0.0751511123445299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,10240,0.05951378080579969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,12288,0.06000177727805244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,8192,0.04978222317165799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,7168,0.03681777914365133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,10240,0.052389333645502724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,8192,0.04549422197871738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,6144,0.032368001010682844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,5120,0.027655111418830022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,7168,0.03791555431154039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,6144,0.03630044394069248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,4096,0.02273066673013899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,5120,0.03259822063975864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,4096,0.02679644525051117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,3072,0.0181040002240075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,3584,0.025768889321221247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,2560,0.015240889456537036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,3072,0.022587555978033278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,2048,0.013236444857385425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,2560,0.021888888544506494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,2048,0.019517333971129525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,1536,0.010527110762066312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,1536,0.017840888765123155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,1024,0.0083155557513237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,3584,0.02017600006527371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,768,0.006934222247865465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,1024,0.015391111373901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,512,0.004963555683692296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,768,0.013756444056828817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,256,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,512,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,128,0.0029288888391521242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,256,0.011710222396585675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,64,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,128,0.011119999819331698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,16384,32,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,64,0.011720889144473605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,16384,32,0.010397333237859938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,65536,0.24477243423461914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,65536,0.15398400359683567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,51200,0.1891626649432712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,16384,0.07301511367162068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,51200,0.11937956015268962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,12288,0.051170667012532554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,16384,0.06608177555931939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,10240,0.044399112462997437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,12288,0.04903022117084927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,8192,0.03726844324005975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,10240,0.044580443037880786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,8192,0.041617777612474226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,7168,0.03293155630429586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,6144,0.028660443094041612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,7168,0.03480799992879232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,5120,0.02546311087078518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,6144,0.030192888445324365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,4096,0.020921778347757127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,5120,0.028732445504930284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,3584,0.01923733287387424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,4096,0.024615110622511968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,3072,0.017133333616786532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,3584,0.023045332895384893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,2560,0.015098666151364645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,3072,0.020591111646758187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,2048,0.010632889138327705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,2560,0.0200284438000785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,1536,0.00884622252649731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,2048,0.018271999226676095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,1024,0.00693866651919153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,1536,0.016545777519543965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,768,0.006176888942718506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,1024,0.013766222529941134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,512,0.003551111039188173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,768,0.01314933349688848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,512,0.01239822225438224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,128,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,256,0.01148088855875863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,64,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,128,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,12288,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,64,0.011009777585665384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,12288,32,0.010808000134097205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,65536,0.20373955037858751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,51200,0.16183466381496853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,65536,0.14777244461907282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,16384,0.06360889143413968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,51200,0.11539378431108262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,16384,0.05465777715047201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,12288,0.04376177655325996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,10240,0.03783199853367276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,8192,0.031007111072540283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,12288,0.042605333858066134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,7168,0.02859555681546529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,10240,0.04086933202213711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,8192,0.03405155407057868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,7168,0.02919911013709174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,6144,0.025682666235499915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,5120,0.02252888845072852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,4096,0.01829333272245195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,6144,0.026956443985303242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,5120,0.024756444825066462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,3584,0.017335111896197002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,4096,0.02143999934196472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,3072,0.015819556183285184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,2560,0.01147377739350001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,3584,0.021327111456129286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,3072,0.019254222512245178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,2048,0.010306666294733683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,1536,0.008747555315494537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,2560,0.017641777793566387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,1024,0.006333333336644703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,1536,0.015484443969196744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,768,0.005577777822812398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,1024,0.012887111140622033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,512,0.004590222405062782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,768,0.01314755529165268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,256,0.003354666754603386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,512,0.012059555285506778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,128,0.002961777771512667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,256,0.012496888637542725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,128,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,2048,0.01682133310370975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,64,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,10240,32,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,64,0.010795555180973478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,10240,32,0.011363555159833698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,65536,0.1642906665802002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,51200,0.1305840015411377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,65536,0.09989244408077663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,16384,0.050176001257366605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,51200,0.07703555292553373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,16384,0.04453955425156487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,12288,0.03494577937655979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,10240,0.02809777855873108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,10240,0.032960888412263654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,8192,0.023084445132149592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,8192,0.026675555441114638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,12288,0.03261244297027588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,7168,0.020655999581019085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,6144,0.018447111050287884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,7168,0.024111110303137038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,5120,0.01606666710641649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,6144,0.02273511058754391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,4096,0.013199110825856527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,5120,0.021485333641370136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,3584,0.011947555674446953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,4096,0.019623999794324238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,3584,0.018547554810841877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,3072,0.010967111421955956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,2560,0.009662222531106737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,3072,0.017127111554145813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,2048,0.008345777789751688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,2560,0.016541333662139047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,1536,0.007326222128338284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,2048,0.015263999501864115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,1024,0.005390222287840313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,1536,0.014116444521480136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,768,0.003923555629120933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,1024,0.012710221939616732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,512,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,768,0.01256622208489312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,512,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,256,0.0032088889016045462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,256,0.011109333071443768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,64,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,128,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,8192,32,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,64,0.01111822244193819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,8192,32,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,65536,0.14394577344258627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,51200,0.11601866616143121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,65536,0.08962755733066136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,16384,0.04148355457517836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,51200,0.07055199808544584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,12288,0.031403554810418025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,16384,0.044181333647833936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,10240,0.027238221632109746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,12288,0.03385511040687561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,8192,0.023052444060643513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,10240,0.03214044372240702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,7168,0.02051555613676707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,7168,0.023572444915771484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,6144,0.016622222132152982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,6144,0.022646221849653456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,5120,0.014848000473446317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,5120,0.02092711130777995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,4096,0.012198222180207571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,8192,0.026146666871176824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,3584,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,4096,0.019368888603316415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,3072,0.010293333066834344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,3584,0.01830933325820499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,2560,0.009364444348547194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,3072,0.017495110630989075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,2048,0.00810222244924969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,2560,0.01607822212908003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,1536,0.006992888947327931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,2048,0.014877332581414117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,1024,0.004234666625658671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,1536,0.013481777575280933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,768,0.003875555677546395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,1024,0.012758221891191272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,512,0.0032675556010670135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,768,0.01183555523554484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,256,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,512,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,256,0.011007111105653973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,128,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,64,0.002552888873550627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,128,0.01072266697883606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,7168,32,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,64,0.011719110939237805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,7168,32,0.009711110757456886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,65536,0.1252266698413425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,51200,0.10104799932903713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,65536,0.08095822069380018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,16384,0.03991733325852288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,51200,0.06475022104051378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,16384,0.04047911034690009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,12288,0.0292871097723643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,12288,0.030228445927302044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,10240,0.024719999896155462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,10240,0.029465778006447688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,8192,0.020389333367347717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,8192,0.026245334082179602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,7168,0.01867822143766615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,7168,0.022691556149058874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,6144,0.016742222838931613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,5120,0.014944000376595391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,6144,0.02137600051032172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,4096,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,5120,0.02037333283159468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,3584,0.010609777437316047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,4096,0.018263111511866253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,3072,0.009415999882751042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,3584,0.01702933344576094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,3072,0.016650666793187458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,2560,0.0155404441886478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,2048,0.007612444460391998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,2048,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,1536,0.006604444649484422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,1536,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,1024,0.003986666599909465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,1024,0.012416889270146688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,768,0.0038515557017591265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,2560,0.008468444148699442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,512,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,768,0.011779555843936073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,256,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,512,0.011716444459226398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,128,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,256,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,64,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,128,0.011386666860845355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,6144,32,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,64,0.010767111347781287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,6144,32,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,65536,0.10540178087022568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,51200,0.08369688855277167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,65536,0.07341511381997003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,16384,0.034703999757766724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,51200,0.05889422363705105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,12288,0.025750221477614507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,16384,0.037323554356892906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,10240,0.021977777282396953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,12288,0.028696888022952612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,10240,0.02719377809100681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,8192,0.01904177831278907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,7168,0.01688088807794783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,8192,0.02369600037733714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,6144,0.015204444527626038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,7168,0.021351999706692163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,5120,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,6144,0.02029511166943444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,4096,0.010422222316265106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,5120,0.019282667173279654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,3584,0.009738666315873465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,4096,0.01759199963675605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,3072,0.008647999829716152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,3584,0.016574222180578444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,2560,0.008184000021881526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,3072,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,2048,0.006959999601046245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,2560,0.014513777361975776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,1536,0.004967999955018361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,2048,0.013771555489963956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,1024,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,1536,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,768,0.0037253333462609183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,1024,0.012376888758606382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,512,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,512,0.011504000259770287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,256,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,256,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,128,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,128,0.011027555498811932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,64,0.0028213332924577924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,64,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,5120,32,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,768,0.011685333318180509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,5120,32,0.010019555687904358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,65536,0.08527200089560615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,51200,0.06800711154937744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,65536,0.06709244516160753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,16384,0.027956446011861164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,51200,0.05442400111092461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,12288,0.021028444170951843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,16384,0.033245331711239286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,10240,0.018518222702874076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,12288,0.02679288884003957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,8192,0.015538665983412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,10240,0.024683554967244465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,7168,0.014166222678290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,8192,0.023006222314304773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,6144,0.013225778109497495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,7168,0.020022221737437777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,5120,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,6144,0.019241778386963736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,5120,0.018546667363908555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,4096,0.010798222488827176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,4096,0.01682577696111467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,3584,0.010943111446168689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,3584,0.01477599971824222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,3072,0.009611555271678502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,3072,0.014091556270917257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,2560,0.008636444807052612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,2560,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,2048,0.007291555404663086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,2048,0.01309511148267322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,1536,0.00591555568906996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,1536,0.01275644451379776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,1024,0.004896000027656555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,1024,0.01204177737236023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,768,0.0041964443193541635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,768,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,512,0.0038480001191298165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,512,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,256,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,256,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,128,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,128,0.011064888702498542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,64,0.0025333333760499954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,64,0.011023110813564725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,4096,32,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,4096,32,0.0099982221921285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,65536,0.0752284460597568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,65536,0.06427644358740912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,51200,0.06047911114162869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,51200,0.052320887645085655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,16384,0.023894222246276006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,16384,0.03124177787039015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,12288,0.018312000566058688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,12288,0.026032888227038916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,10240,0.01587733295228746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,10240,0.023553777072164748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,8192,0.014053333136770459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,8192,0.02106311089462704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,7168,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,7168,0.019235556324323017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,6144,0.012273777690198688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,6144,0.0185244447655148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,5120,0.011375111010339526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,5120,0.017679111825095285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,4096,0.010496888723638324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,3584,0.00996888925631841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,4096,0.015527110960748462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,3072,0.009375111096435124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,3584,0.014820444915029736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,2560,0.007992000215583378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,3072,0.013774221969975365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,2048,0.00676711110605134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,2560,0.013724444641007317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,2048,0.013168888787428537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,1536,0.005569777968857024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,1024,0.004620444443490771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,1536,0.012662221988042196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,768,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,1024,0.01147911118136512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,512,0.004577777865860197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,768,0.011760888828171624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,256,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,512,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,256,0.011557333171367645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,64,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,128,0.011132444772455426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,64,0.011071110765139261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3584,32,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,65536,0.06514666477839152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3584,32,0.00977422214216656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,51200,0.050640000237358936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,65536,0.06136888927883572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,16384,0.020946666598320007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,51200,0.050092445479498975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,12288,0.016550223032633465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,16384,0.030464887619018555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,10240,0.014895111322402954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,12288,0.02458311120669047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,8192,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,10240,0.022799111074871484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,8192,0.020033778415785897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,7168,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,6144,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,7168,0.018831110662884183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,5120,0.010108444425794814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,6144,0.018016889691352844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,4096,0.009760888914267222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,5120,0.016541333662139047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,3584,0.009017777939637503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,4096,0.014838222000333997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,3072,0.008654221892356873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,3584,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,2560,0.007116444408893585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,3072,0.013441777891582914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,2048,0.006161777923504512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,2560,0.013440888788965015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,1536,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,2048,0.01276000009642707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,1536,0.01240888900227017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,1024,0.004993777722120285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,1024,0.011784888803958893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,768,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,768,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,512,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,512,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,256,0.0035031110876136353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,256,0.011501332951916588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,128,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,128,0.011037333144081963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,64,0.002505777817633417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,64,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,3072,32,0.002492444382773505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,3072,32,0.009717333647939894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,65536,0.05540800094604492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,65536,0.05925066603554619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,51200,0.04519199993875292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,51200,0.048707554737726845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,16384,0.020000888241661918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,16384,0.027148443791601393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,12288,0.015072888798183866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,12288,0.023399111297395494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,10240,0.01384800010257297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,10240,0.020983111527231004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,8192,0.011961778004964193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,8192,0.01926399932967292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,7168,0.011174221833546957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,7168,0.018211555149820115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,6144,0.010633777413103314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,6144,0.016593777471118502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,5120,0.009714666340086196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,5120,0.01462311049302419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,4096,0.009342222577995723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,4096,0.014185777968830533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,3584,0.00832799987660514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,3584,0.013782222237851886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,3072,0.007678222325113084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,3072,0.013388444152143268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,2560,0.007091555330488417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,2560,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,2048,0.0058497778243488735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,2048,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,1536,0.005576888720194499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,1536,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,1024,0.004649777793222004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,768,0.004209777961174647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,1024,0.011747555600272285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,512,0.003559111307064692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,768,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,256,0.0032435556252797446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,512,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,128,0.0028204443968004654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,256,0.011735111474990845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,64,0.0028168888141711554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,128,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2560,32,0.002510222295920054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,64,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2560,32,0.009664889011118147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,65536,0.05721600188149346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,51200,0.037372443411085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,51200,0.04733777708477444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,16384,0.015063110325071545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,16384,0.024471999870406255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,12288,0.012539555629094442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,65536,0.04531822270817227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,12288,0.02100444502300686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,10240,0.011704000333944956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,8192,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,10240,0.01958577831586202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,7168,0.010448888772063784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,7168,0.01626311077011956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,6144,0.009835555321640438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,8192,0.018332444959216647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,5120,0.009011555876996782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,4096,0.008359111017651027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,6144,0.015214223000738354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,3584,0.007373332977294922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,4096,0.013286222186353473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,3072,0.006304889089531369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,3584,0.01309333327743742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,5120,0.014269333746698169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,2560,0.005551110953092575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,2048,0.007333333293596904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,3072,0.012864888542228274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,2560,0.012699555191728802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,1536,0.005568000177542369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,2048,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,1024,0.004598222259018156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,768,0.004217777815130022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,1536,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,1024,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,512,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,768,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,256,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,128,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,512,0.01148888882663515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,64,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,256,0.011038222246699862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,2048,32,0.002805333377586471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,128,0.011424889167149862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,64,0.011042666931947073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,65536,0.035875555541780256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,2048,32,0.010695999695195092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,51200,0.028805332051383123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,65536,0.055535111162397593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,16384,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,51200,0.045758220884535045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,12288,0.011130666567219628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,16384,0.02300888962215847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,10240,0.013602666556835175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,12288,0.020590222544140287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,10240,0.018227555685573153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,8192,0.01182755579551061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,7168,0.011524444652928246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,8192,0.016344888342751395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,6144,0.008635555704434713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,7168,0.015508444772826301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,5120,0.007679999702506595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,6144,0.014853333433469137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,4096,0.007237333390447829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,5120,0.014120888378885059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,3584,0.007318221860461765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,4096,0.013411555853154926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,3072,0.006241777704821692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,3584,0.013541333377361298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,2560,0.005650666852792104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,3072,0.013099555340078143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,2048,0.005752888818581899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,2560,0.012476444244384766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,1536,0.0052151112920708126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,2048,0.012192000117566852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,1024,0.00424888895617591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,1536,0.012121777567598553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,1024,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,768,0.01146399974822998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,512,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,512,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,256,0.003248888792263137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,256,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,128,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,128,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,768,0.003920000046491623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,64,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1536,32,0.00254844439526399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,64,0.0114720000161065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1536,32,0.009755555954244401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,65536,0.025404444999165002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,51200,0.020945777495702107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,65536,0.05240977803866068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,16384,0.01420177850458357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,16384,0.021764443980322942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,51200,0.04403822289572822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,12288,0.018236445056067575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,10240,0.010973333484596677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,10240,0.016905777984195285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,8192,0.010068444742096795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,8192,0.015860444969601102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,7168,0.009391999906963771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,12288,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,6144,0.0083137775460879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,7168,0.015573332707087198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,5120,0.007364444434642792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,6144,0.014819555812411837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,4096,0.0069875551594628235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,5120,0.014112888111008538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,3584,0.006643555644485686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,4096,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,3072,0.006099555641412735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,3584,0.01348444405529234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,3072,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,2560,0.005623999983072281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,2048,0.006255110932721033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,2560,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,1536,0.00527555536892679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,2048,0.012396444049146442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,1024,0.004294222427739037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,1536,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,768,0.003939555750952827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,1024,0.01183644433816274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,512,0.003559999995761447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,768,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,256,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,512,0.01179377817445331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,256,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,128,0.003055111194650332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,64,0.0028115556471877625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,1024,32,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,128,0.011069333387745751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,65536,0.021496888664033677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,32,0.010690666735172272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,51200,0.01779733267095354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,65536,0.05141777793566386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,1024,64,0.010924444430404238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,51200,0.043142222695880465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,12288,0.010449777874681683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,16384,0.021661332911915247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,12288,0.01811911165714264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,10240,0.00944977750380834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,8192,0.009133333133326637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,10240,0.017112889223628573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,8192,0.01593777702914344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,7168,0.008440888590282863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,16384,0.012418666647540199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,6144,0.006924444602595435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,7168,0.015537778536478678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,5120,0.00664444433318244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,6144,0.014819555812411837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,4096,0.007346666521496243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,5120,0.014132445057233175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,3584,0.0069724445541699724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,4096,0.013533333109484779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,3072,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,3584,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,2560,0.006236444330877728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,3072,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,2048,0.00489777781897121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,2560,0.012720000412729053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,1536,0.004911111046870549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,2048,0.012153777811262341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,1024,0.003907555507289039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,1024,0.011403555671374003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,768,0.0038648889296584656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,1536,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,512,0.003452444449067116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,768,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,512,0.01204711116022534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,256,0.0114720000161065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,64,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,128,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,768,32,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,64,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,65536,0.02239466706911723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,65536,0.050958222813076444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,51200,0.018725333942307364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,768,32,0.010774222513039907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,16384,0.010423111418883005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,51200,0.04230044285456339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,12288,0.008731555607583787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,16384,0.020371556282043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,10240,0.007684444387753804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,12288,0.01789155602455139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,8192,0.00797155582242542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,10240,0.016897777716318767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,7168,0.007654222349325816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,8192,0.015790222419632804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,6144,0.007292444507280986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,7168,0.015480000111791821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,5120,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,6144,0.014682667122946845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,4096,0.007276444799370236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,5120,0.014147554834683737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,3584,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,4096,0.013418667018413544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,3072,0.006583111153708563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,3584,0.01351377781894472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,2560,0.00658044425977601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,3072,0.012746666868527731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,2048,0.005952888892756567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,2560,0.012618666721714867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,1536,0.004556444370084339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,2048,0.012111999922328524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,1024,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,1536,0.012147555748621622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,768,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,768,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,512,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,512,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,256,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,256,0.011775111158688864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,128,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,128,0.011405333048767514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,64,0.002833777831660377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,1024,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,512,32,0.002571555475393931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,64,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,512,32,0.009696000152164036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,65536,0.014492443866199918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,51200,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,65536,0.048935112026002675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,16384,0.008641777767075432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,51200,0.03975111246109009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,12288,0.00832177781396442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,16384,0.019987555013762582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,10240,0.007368000017272101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,12288,0.018216888109842937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,8192,0.0069715554515520734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,10240,0.01686044368478987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,7168,0.006581333362393909
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,8192,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,6144,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,7168,0.015135111080275642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,5120,0.005951111101441913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,6144,0.01455200049612257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,4096,0.006596444381607904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,5120,0.014031110538376702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,3584,0.006250666661394968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,4096,0.013409777647919126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,3584,0.013069333301650153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,3072,0.006009777800904379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,3072,0.01274222218328052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,2560,0.006227555374304454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,2560,0.012791111237472959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,2048,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,2048,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,1536,0.004540444248252445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,1536,0.012067555553383298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,1024,0.0036720000207424164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,1024,0.01144177797767851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,768,0.003436444534195794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,768,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,512,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,512,0.011468444433477191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,256,0.003176888864901331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,256,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,128,0.0029688889367712867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,64,0.0025742221623659134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,128,0.011310222248236338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,256,32,0.002491555487116178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,64,0.010129777921570672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,65536,0.01351199961370892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,256,32,0.009673777553770278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,51200,0.01184177812602785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,65536,0.04591644472546048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,16384,0.00758844448460473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,51200,0.03906844390763177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,12288,0.007273777491516537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,16384,0.02012355625629425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,12288,0.01805688937505086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,10240,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,8192,0.00665866666369968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,10240,0.01684533390733931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,7168,0.006596444381607904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,8192,0.01518133282661438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,6144,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,7168,0.014800889624489678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,5120,0.006285333385070165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,6144,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,4096,0.005543111099137201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,5120,0.013803555733627744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,3584,0.006269333263238271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,4096,0.01276622215906779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,3072,0.005525333185990651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,3584,0.013076444466908773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,2560,0.00619555554456181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,3072,0.012426666915416718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,2048,0.0052560000783867305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,2560,0.012783110969596438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,1536,0.004510222209824456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,2048,0.011448889142937131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,1024,0.0037155557009908888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,1536,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,768,0.0035253332720862497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,1024,0.011009777585665384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,512,0.003035555490189128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,768,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,256,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,512,0.010987555815113915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,128,0.002564444517095884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,256,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,64,0.0024888888001441956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,128,0.01034311122364468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,128,32,0.002542222125662698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,64,0.011078221930397881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,128,32,0.008941333327028487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,65536,0.045929776297675244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,51200,0.009727111293209923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,51200,0.03806311223242018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,16384,0.006928000185224745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,16384,0.01960444450378418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,12288,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,12288,0.01754844519827101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,10240,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,65536,0.01071911139620675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,8192,0.006912888752089606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,10240,0.016532444291644625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,7168,0.005946666830115848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,8192,0.015487111277050443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,6144,0.006269333263238271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,7168,0.014835556348164877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,5120,0.005559111220969095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,6144,0.0144533332851198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,4096,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,5120,0.013112000293201871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,3584,0.005606222069925732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,4096,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,3072,0.005592000153329637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,3584,0.012726222475369772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,2560,0.005582222094138463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,3072,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,2048,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,2560,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,1536,0.004210666649871402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,2048,0.012078222301271228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,1024,0.003879111260175705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,1536,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,768,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,1024,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,768,0.011023110813564725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,512,0.003132444495956103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,512,0.011022222538789114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,256,0.0027742222365405825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,256,0.0107004443804423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,128,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,128,0.01033155537313885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,64,0.0028124445428450904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,64,0.010054222411579555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,64,32,0.002777777819169892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,64,32,0.009017777939637503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,65536,0.009100444614887238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,51200,0.008244444098737504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,65536,0.04564088914129469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,16384,0.006250666661394968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,16384,0.019725332657496136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,12288,0.005960000058015187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,51200,0.0389013323518965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,10240,0.006574222197135289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,10240,0.016709332664807636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,8192,0.005743999862008625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,8192,0.015676443775494892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,7168,0.005700444595681296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,12288,0.017498667041460674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,6144,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,7168,0.01512977812025282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,5120,0.0052328887912962176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,6144,0.014171555638313293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,4096,0.0053182223604785065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,4096,0.012784889174832238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,5120,0.013704000247849358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,3584,0.005640888793600931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,3072,0.0052915554907586836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,3584,0.01312622262371911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,3072,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,2560,0.005712888720962737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,2048,0.005413333161009683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,2560,0.012459555433856117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,1536,0.0041653331783082755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,2048,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,1024,0.0037191112836201987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,1536,0.01146666705608368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,768,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,1024,0.0107022225856781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,768,0.01076533314254549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,512,0.0032133333798911837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,256,0.002564444517095884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,512,0.01109866632355584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,128,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,256,0.01071111112833023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,128,0.010277333358923594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,64,0.002583111118939188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,7,32,32,0.002499555548032125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,64,0.010033778018421596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,7,32,32,0.00999111102686988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,51200,0.49236445956759983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,51200,0.9668898052639432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,16384,0.3322524494594998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,16384,0.24821866883171928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,12288,0.24625955687628853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,10240,0.20739289124806723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,12288,0.15594222810533312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,8192,0.16928799947102866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,10240,0.1621102227105035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,8192,0.13788089487287733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,7168,0.15012266900804308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,7168,0.10320622391170925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,6144,0.11522577868567573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,5120,0.0976746678352356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,6144,0.09218044413460626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,4096,0.07732088698281182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,5120,0.09506933556662665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,4096,0.06623288657930163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,3584,0.06877599822150336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,3072,0.05917155742645264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,3584,0.06869777705934313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,2560,0.050329777929517955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,3072,0.05205688873926798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,2048,0.041174223025639854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,2560,0.04803555541568332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,1536,0.032199111249711775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,2048,0.03835377759403653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,1024,0.02277155551645491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,1536,0.03405422303411696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,768,0.018444443742434185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,1024,0.02622577713595496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,512,0.013451555536852943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,768,0.022592888938056097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,256,0.008931555681758458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,512,0.018534221582942538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,256,0.015442666080262927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,128,0.005906666732496685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,64,0.00553955551650789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,128,0.013155555559529198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,65536,32,0.005655111124118169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,64,0.014888000157144336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,65536,32,0.013233777549531726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,65536,0.5188284450107151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,65536,1.079811520046658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,51200,0.8430675400627984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,51200,0.39880977736579043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,16384,0.2736995485093859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,16384,0.15465688705444336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,12288,0.2000666724310981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,12288,0.16181955072614881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,10240,0.16964711083306205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,10240,0.1301413377126058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,8192,0.13992888397640652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,8192,0.09697777695126003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,7168,0.12275911702050103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,7168,0.08656266662809585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,6144,0.10769066545698379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,6144,0.08345422479841445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,5120,0.09283110830518936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,4096,0.07132089138031006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,5120,0.07299644417232938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,3584,0.061062223381466336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,4096,0.05626577801174588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,3072,0.05336177680227491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,3584,0.05726133452521431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,2560,0.04450311263402303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,3072,0.04539377821816338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,2048,0.03770222266515096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,2560,0.04185600082079569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,1536,0.02885955572128296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,2048,0.03360088997417026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,1024,0.02121066715982225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,1536,0.030096001095241968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,768,0.01589777734544542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,1024,0.023286221755875483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,768,0.021013332737816706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,512,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,256,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,512,0.01773066653145684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,128,0.003916444463862313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,256,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,64,0.003212444484233856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,128,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,51200,32,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,64,0.013815110756291283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,51200,32,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,65536,0.3282151222229004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,51200,0.2576826678382026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,65536,0.18052621682484946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,16384,0.09532711240980361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,12288,0.06875289148754544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,51200,0.13839733600616455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,16384,0.07467555337482028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,10240,0.05953511264589098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,8192,0.05001066790686714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,12288,0.06216000186072456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,10240,0.05506844321886698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,7168,0.037018666664759316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,6144,0.03235377868016561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,8192,0.043578667773140796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,5120,0.02788089050187005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,7168,0.04070666763517592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,4096,0.023004444109068975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,6144,0.03785422113206651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,3584,0.020584889584117465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,5120,0.034889777501424156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,3072,0.01795733306143019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,4096,0.02633511192268795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,3584,0.026536888546413843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,2560,0.015740444262822468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,2048,0.013167111410035027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,3072,0.023092443744341534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,2560,0.022090666823916968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,1536,0.010402667025725046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,2048,0.019097778532240126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,1024,0.008150222400824228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,768,0.0069448889957533936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,1024,0.015357333752844067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,512,0.004580444345871608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,1536,0.0188417782386144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,768,0.013821333646774292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,256,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,128,0.0032248888164758682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,512,0.01350933313369751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,64,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,256,0.011577777564525604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,16384,32,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,128,0.012135999898115793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,64,0.01146933353609509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,65536,0.24205954869588217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,16384,32,0.011043555206722684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,65536,0.1533164448208279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,51200,0.18876977761586508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,16384,0.07306222120920818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,12288,0.05095555716090732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,51200,0.11979377269744873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,10240,0.04415200153986613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,16384,0.06223022275500827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,12288,0.054619554016325206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,8192,0.03713333275583055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,10240,0.048178666167789035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,7168,0.03287911083963182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,8192,0.038387556870778404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,6144,0.028375112348132666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,5120,0.02555999987655216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,6144,0.033041778537962176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,7168,0.03433955709139506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,4096,0.020993777447276648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,5120,0.029691557089487713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,3584,0.018939556346999276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,4096,0.024454222785101995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,3072,0.017087999317381117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,3584,0.02221244408024682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,2560,0.015219555960761176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,3072,0.021313778228229944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,2048,0.010780444575680627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,2560,0.0203493336836497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,1536,0.008678221868144141
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,1536,0.01661333276165856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,2048,0.017290666699409485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,1024,0.006966222491529252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,768,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,1024,0.013849777479966482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,512,0.0037946667936113146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,768,0.013832888669437833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,256,0.0032053333189752367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,512,0.011834666960769229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,256,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,64,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,128,0.011503111157152386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,12288,32,0.0028897778441508612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,64,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,12288,32,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,65536,0.2029057741165161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,65536,0.14696621894836426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,51200,0.16208799680074057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,16384,0.06291911337110731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,51200,0.11547199885050456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,12288,0.04348444607522753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,16384,0.05401244428422716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,12288,0.04680533210436503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,10240,0.03769955701298184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,8192,0.030996445152494643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,10240,0.03942755526966519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,7168,0.02866577770974901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,8192,0.03253688746028476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,6144,0.025427555044492085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,7168,0.030123554997973975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,5120,0.022065778573354084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,6144,0.027585777971479628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,5120,0.024863999750879075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,4096,0.018114666144053142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,3584,0.017224889662530687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,4096,0.021664889322386846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,3072,0.015604444675975375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,3584,0.01963911122745938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,2560,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,3072,0.019596444235907663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,2048,0.01035911093155543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,2560,0.017920888132519193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,1536,0.008778666456540426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,2048,0.01688799924320645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,1024,0.006736889067623351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,1536,0.014735110931926303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,768,0.0053173332578606075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,1024,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,512,0.00423555572827657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,768,0.01278133359220293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,512,0.012152888708644442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,128,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,256,0.011238222320874533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,64,0.00250133333934678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,64,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,10240,32,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,32,0.010684444672531553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,65536,0.1647475560506185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,10240,128,0.011400889191362592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,65536,0.097853336069319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,51200,0.13076533211602107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,16384,0.04979733294910855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,51200,0.07699733310275607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,12288,0.0328382220533159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,16384,0.04271289043956333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,10240,0.028354667954974707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,12288,0.03844622108671401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,8192,0.023394667439990576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,10240,0.032356444332334734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,7168,0.020784000555674236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,8192,0.02607288956642151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,7168,0.0247013337082333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,6144,0.01854222185081906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,5120,0.015826667348543804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,6144,0.023356444305843774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,4096,0.013075555364290873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,5120,0.02164799968401591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,3584,0.011929777761300405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,4096,0.018911111685964797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,3072,0.010810666614108615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,3584,0.017356443736288283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,2560,0.009663999908500248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,3072,0.017258667283587985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,2048,0.008361777497662438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,2560,0.016176000237464905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,2048,0.014779556128713818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,1536,0.0069573331210348345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,1536,0.013781333135233985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,1024,0.005229333208666907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,1024,0.012783110969596438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,768,0.003625777860482534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,512,0.003649777836269803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,768,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,256,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,512,0.01072266697883606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,128,0.0029279999434947968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,256,0.011648889217111798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,64,0.0026151111556424033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,128,0.010672888822025724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,8192,32,0.0029022221763928733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,64,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,8192,32,0.009767110976907942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,65536,0.14473156134287515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,51200,0.11625866095225017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,65536,0.08970933490329319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,16384,0.04227644536230299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,51200,0.07075911098056369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,12288,0.031367109881507024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,16384,0.04502844479348925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,10240,0.027493332823117573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,12288,0.03829955392413669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,8192,0.02272355556488037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,10240,0.030518223841985066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,7168,0.020544000797801547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,8192,0.026110221942265827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,6144,0.01680888897842831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,7168,0.02442755632930332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,5120,0.014745778507656522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,6144,0.02269777821169959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,4096,0.012356444365448423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,5120,0.020256888535287645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,3584,0.011297778122954898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,4096,0.01854044364558326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,3072,0.01001422190003925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,3584,0.017909333109855652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,2560,0.00905955582857132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,3072,0.017315554949972365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,2048,0.007857778006129796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,2560,0.015864888827006023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,1536,0.006681777950790193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,2048,0.01386222243309021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,1024,0.0041848888827694785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,1536,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,768,0.003891555385457145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,1024,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,512,0.00332622230052948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,768,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,256,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,512,0.011712888876597086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,128,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,256,0.01110755569405026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,64,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,128,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,7168,32,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,64,0.01000355515215132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,7168,32,0.010120000276300643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,65536,0.1250524388419257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,51200,0.10082044204076131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,65536,0.08091821935441759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,16384,0.03983822133806016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,51200,0.06431644492679171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,12288,0.02909066610866123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,16384,0.03727733426623874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,10240,0.024694222542974684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,12288,0.03105688757366604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,8192,0.02053955528471205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,10240,0.029531555043326482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,7168,0.018926221463415358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,8192,0.026367111338509455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,6144,0.01681511104106903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,7168,0.023186667097939387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,5120,0.015080889066060385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,6144,0.021560889151361253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,4096,0.011305777562989129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,5120,0.02134399943881565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,3584,0.010693333215183683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,4096,0.017913776967260573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,3072,0.00960000024901496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,3584,0.017112889223628573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,2560,0.008471999731328752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,3072,0.016194666425387066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,2048,0.007611555357774098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,2560,0.015657777587572735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,1536,0.00629333323902554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,2048,0.013616888887352414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,1024,0.004129777765936322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,1536,0.013154666456911298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,768,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,1024,0.012090666426552666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,512,0.0035208890007601846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,768,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,256,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,512,0.010671999719407825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,128,0.0029191111938820947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,256,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,128,0.010413332945770687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,64,0.0029208889851967492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,64,0.010063110954231685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,6144,32,0.0028648889727062653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,65536,0.10515466663572524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,65536,0.07291199763615926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,6144,32,0.011113777756690979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,51200,0.08367821905348037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,16384,0.034634666310416326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,51200,0.059296886126200356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,12288,0.02574933403068119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,16384,0.03915289044380188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,10240,0.021670222282409668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,12288,0.03155022197299533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,8192,0.019102222389645047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,10240,0.02809244394302368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,7168,0.016872889465755887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,8192,0.023524444964196947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,7168,0.02201066745652093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,6144,0.015209777487648858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,5120,0.013731555806265937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,6144,0.020815110868877836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,4096,0.010402667025725046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,5120,0.020100444555282593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,3584,0.009722666607962714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,4096,0.017208000024159748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,3072,0.008816888762844933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,3584,0.0169840008020401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,2560,0.008185777399275038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,3072,0.014509333504570855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,2048,0.007167999943097432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,2048,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,1536,0.00519466648499171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,1536,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,1024,0.003995555556482739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,1024,0.011815999945004782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,768,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,768,0.012206222448084088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,512,0.003530666646030214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,2560,0.014871110518773397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,512,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,256,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,128,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,256,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,128,0.01145155562294854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,5120,32,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,64,0.010007999837398529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,65536,0.08577066659927368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,5120,32,0.010750222537252637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,51200,0.06800888644324408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,65536,0.06714310910966662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,16384,0.02771111163828108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,51200,0.0544435547457801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,16384,0.03513600097762214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,12288,0.021020443903075323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,10240,0.018633777896563213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,12288,0.02685600022474925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,8192,0.015554666519165039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,10240,0.026808000273174707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,7168,0.014487110906177096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,8192,0.02273955610063341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,6144,0.013020444247457715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,7168,0.02096533278624217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,5120,0.012019555601808759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,6144,0.01964000033007728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,4096,0.01071999967098236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,5120,0.019162666466501024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,4096,0.016140444411171805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,3584,0.01073066641887029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,3072,0.009696000152164036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,3584,0.015551111764378019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,2560,0.008671110702885522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,3072,0.014522666732470194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,2048,0.007431999676757389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,2560,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,1536,0.005923555543025334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,2048,0.013584888643688627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,1536,0.012098666694429187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,1024,0.005715555614895291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,768,0.004596444467703502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,1024,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,512,0.003909333298603694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,768,0.012436444560686747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,256,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,512,0.010431999961535135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,128,0.0029226665695508323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,256,0.01144444445768992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,128,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,64,0.002528000002106031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,4096,32,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,64,0.009759110709031424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,4096,32,0.0107342220014996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,65536,0.07526399691899617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,51200,0.06039200226465861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,65536,0.06413066387176514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,16384,0.023749333288934495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,51200,0.052095999320348106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,12288,0.018269333574506972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,16384,0.030618667602539062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,12288,0.025768889321221247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,10240,0.01626844373014238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,10240,0.024510223004553053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,8192,0.01421333352724711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,8192,0.02187999917401208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,7168,0.013463999662134381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,7168,0.019673777951134574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,6144,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,6144,0.01866311166021559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,5120,0.01125511113140318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,5120,0.01721244388156467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,4096,0.010377777947319878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,4096,0.015494222442309061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,3584,0.01014311114947001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,3584,0.014382221632533602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,3072,0.00941066692272822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,3072,0.01348533315791024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,2560,0.008344888687133789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,2560,0.013439111411571503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,2048,0.006760889043410619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,2048,0.012454222473833295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,1536,0.005570666657553778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,1536,0.012108444339699216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,1024,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,1024,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,768,0.00443822228246265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,768,0.0118568887313207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,512,0.005613333235184352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,512,0.010407110883129967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,256,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,256,0.011029333704047732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,128,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,128,0.010334221853150262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,64,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,64,0.009720889230569204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3584,32,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3584,32,0.009301332963837517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,65536,0.06530577606625028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,65536,0.06065689192877876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,51200,0.05065066615740458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,51200,0.05002311203214857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,16384,0.02104355560408698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,16384,0.03151022063361274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,12288,0.016275554895401
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,12288,0.024863110648261175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,10240,0.015119110544522604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,10240,0.023688889212078516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,8192,0.013077333569526672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,8192,0.02134399943881565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,7168,0.0121760004096561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,7168,0.019136000010702345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,6144,0.010978666444619497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,6144,0.018070222602950204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,5120,0.01035288886891471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,5120,0.016526222229003906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,4096,0.009776888622177972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,4096,0.014543111125628153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,3584,0.009172444542249044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,3584,0.01404088901148902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,3072,0.008428444465001425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,3072,0.013071999781661563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,2560,0.007323555648326874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,2560,0.01310488912794325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,2048,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,2048,0.012113778127564324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,1536,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,1536,0.012056888805495368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,1024,0.004584889031118817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,1024,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,768,0.004534222185611725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,768,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,512,0.004260444392760594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,512,0.010344888601038191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,256,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,128,0.002959111084540685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,256,0.01073155552148819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,64,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,128,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,64,0.009736888938479954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,3072,32,0.0025840000145965153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,65536,0.05502133236991035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,3072,32,0.009296000003814697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,51200,0.04494933287302653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,65536,0.05868622329499987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,16384,0.020112888680564035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,51200,0.05033066537645128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,12288,0.015944000747468736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,10240,0.01425244410832723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,16384,0.02885155545340644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,8192,0.012166221936543783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,12288,0.0244151105483373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,7168,0.011201777391963534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,10240,0.0229751103454166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,6144,0.010951999988820819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,8192,0.02033333314789666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,5120,0.009654222263230218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,7168,0.019036443697081674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,6144,0.01755911111831665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,4096,0.008970666262838576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,3584,0.008291555775536431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,5120,0.014968888627158271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,4096,0.013879111243618859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,3072,0.007970666719807519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,3584,0.013824889229403602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,2560,0.006604444649484422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,3072,0.013274666335847644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,2048,0.006298666612969504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,2560,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,1536,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,2048,0.011891555454995898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,1024,0.004580444345871608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,1536,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,768,0.0041955556306574075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,1024,0.011161777708265515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,512,0.0035377778112888336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,768,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,256,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,512,0.010448888772063784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,128,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,256,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,64,0.002499555548032125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,128,0.010384889112578498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2560,32,0.0028408887899584244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,64,0.010416000253624385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,65536,0.04498399959670173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2560,32,0.009367999931176504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,51200,0.037626667155159846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,65536,0.05687555339601305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,16384,0.01484266585773892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,51200,0.046915554338031344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,12288,0.012730666332774691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,16384,0.023789332972632513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,10240,0.01148088855875863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,12288,0.02171733313136631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,8192,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,10240,0.020260444945759244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,7168,0.010476444330480365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,8192,0.01883555617597368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,6144,0.00983911090426975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,7168,0.016033778587977093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,5120,0.00868622213602066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,6144,0.0147706667582194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,4096,0.008030222521887885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,5120,0.01420088940196567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,3584,0.007264888948864407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,4096,0.01313066648112403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,3072,0.00628266649113761
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,3584,0.013375110924243927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,2560,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,3072,0.012359999948077731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,2048,0.00665155549844106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,2560,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,1536,0.00555644432703654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,2048,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,1024,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,1536,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,768,0.004211555752489301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,768,0.011710222396585675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,512,0.003560888684458203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,256,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,512,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,128,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,256,0.011387555963463254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,128,0.011329777538776398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,64,0.010444444086816577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,32,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,1024,0.011194667054547204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,2048,32,0.009670221971140968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,65536,0.03568177753024631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,51200,0.028909332222408716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,65536,0.05502133236991035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,16384,0.013262222210566202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,51200,0.044897778166664966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,12288,0.011194667054547204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,16384,0.023265777362717524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,10240,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,2048,64,0.0028026666906144884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,12288,0.020305777589480083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,8192,0.012778666284349231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,7168,0.011321778098742167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,10240,0.01889333294497596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,8192,0.01606044504377577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,6144,0.008646222452322641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,5120,0.007638221813572778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,7168,0.015485333071814643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,6144,0.014490667316648694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,4096,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,5120,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,3584,0.007068444457319047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,3072,0.006248000181383557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,4096,0.013081777426931592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,2560,0.0058551111982928375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,3584,0.013412444127930535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,2048,0.005621333503060871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,2560,0.012473777764373355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,1536,0.004920888692140579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,3072,0.012882666455374824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,1024,0.0042168887125121225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,1536,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,768,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,1024,0.011804444094498953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,768,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,512,0.0035377778112888336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,512,0.011411555939250521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,256,0.0031848889258172777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,2048,0.012469333079126147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,256,0.011145778000354767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,64,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,128,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,64,0.010112889111042023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1536,32,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,65536,0.025035555164019268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1536,32,0.009675555759006077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,51200,0.020717332760492962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,65536,0.05208533340030246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,16384,0.01386844449573093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,51200,0.042958223157458834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,12288,0.011967110964987012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,16384,0.022103110949198406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,10240,0.010750222537252637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,12288,0.018534221582942538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,8192,0.010102222363154093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,10240,0.0169804443915685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,8192,0.015847999188635085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,7168,0.009315555294354757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,6144,0.007271111011505127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,7168,0.0151555554734336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,5120,0.006778666542636023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,6144,0.014862222803963555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,4096,0.00699733363257514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,5120,0.014206222361988492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,3584,0.006519111080302133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,4096,0.013608000344700284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,3584,0.013460444079505073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,3072,0.0063697778516345555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,2560,0.005618666609128316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,3072,0.012840000291665396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,2048,0.006021333237489064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,2560,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,1536,0.005639111002286275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,2048,0.012413333687517377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,1024,0.004231111043029361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,1536,0.012137778103351593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,768,0.004238222208287981
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,1024,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,512,0.003504000190231535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,768,0.011807999677128263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,256,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,512,0.011050666371981302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,128,0.0029226665695508323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,128,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,64,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,256,0.011575111084514193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,1024,32,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,64,0.010416888528399997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,65536,0.021220443977249995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,1024,32,0.010787555740939246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,65536,0.05148266752560934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,51200,0.01771288944615258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,16384,0.01223111069864697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,51200,0.042825778325398765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,12288,0.010556444525718689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,16384,0.020422221885787118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,12288,0.01829955478509267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,10240,0.009418666362762451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,8192,0.008709333009190029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,10240,0.01695999999841054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,7168,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,8192,0.015970667203267414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,7168,0.015246222416559855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,6144,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,5120,0.006637333167923822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,6144,0.014524444937705994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,4096,0.007349333001507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,4096,0.013148444394270578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,5120,0.014262222581439547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,3584,0.00702044450574451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,3072,0.006673777682913675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,3072,0.012865777644846173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,3584,0.01315822203954061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,2560,0.012835555606418185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,2048,0.0053075556125905775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,2048,0.012597333225939008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,1536,0.005008888741334279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,2560,0.006342222293217977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,1024,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,1536,0.012158222496509552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,768,0.003930666794379552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,1024,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,512,0.0038240001433425476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,512,0.01108888867828581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,256,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,768,0.011758222348160215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,256,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,64,0.002749333365095986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,128,0.01146666705608368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,64,0.010410666465759277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,768,32,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,768,32,0.009733333355850643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,65536,0.02199111051029629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,65536,0.05095555716090732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,51200,0.018730666902330186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,51200,0.04233955674701267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,16384,0.010446222292052375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,16384,0.020310221446885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,12288,0.008385777473449707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,12288,0.018186666899257235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,10240,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,10240,0.01700355609258016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,8192,0.008296888735559251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,8192,0.015796444482273526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,7168,0.007657777931955125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,7168,0.015150222513410779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,6144,0.007283555964628856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,6144,0.014495111174053617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,5120,0.006837333242098491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,5120,0.014142221874660917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,4096,0.0073315559162033936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,4096,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,3584,0.007048000064161088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,3584,0.013101333545313941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,3072,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,3072,0.012801777985360889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,2560,0.006254222244024277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,2560,0.012457778056462606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,2048,0.005985777825117111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,2048,0.01240711079703437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,1536,0.0041999999019834726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,1536,0.011721777419249216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,1024,0.003636444608370463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,1024,0.011708444191349877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,768,0.003581333491537306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,768,0.011731555892361535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,512,0.0032915555768542816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,512,0.011030221978823343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,256,0.0031715554909573663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,256,0.011371555427710215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,128,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,128,0.011026666396194033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,64,0.002504000026318762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,64,0.01037333326207267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,512,32,0.002946666752298673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,65536,0.015046222342385186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,512,32,0.009706666900051964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,51200,0.012929778132173749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,51200,0.03949600127008226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,16384,0.008664888640244802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,65536,0.04852622085147434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,16384,0.01998222205373976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,12288,0.007979555262459649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,10240,0.007737778127193451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,12288,0.018240888913472492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,8192,0.006962666908899943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,8192,0.015629332926538255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,7168,0.00666844430896971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,10240,0.016756445169448853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,7168,0.015170666906568738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,6144,0.006147555592987273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,5120,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,6144,0.014574221438831754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,4096,0.006636444479227066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,4096,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,5120,0.013955555028385587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,3584,0.0129439996348487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,3072,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,3072,0.012531555361217923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,2560,0.005662222289376789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,3584,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,2048,0.006053333481152852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,2048,0.012102222277058495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,2560,0.01221244451072481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,1536,0.004271111140648524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,1024,0.00388355553150177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,1024,0.011512888802422417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,768,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,1536,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,512,0.003302222324742211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,768,0.011459555890825061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,256,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,512,0.010682666467295753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,128,0.0028364445186323593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,256,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,64,0.0028764444092909494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,128,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,64,0.010398222340477837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,256,32,0.0024906665914588505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,65536,0.013793778088357715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,51200,0.01186222251918581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,256,32,0.009693333672152625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,16384,0.007303111255168915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,65536,0.04600977897644043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,12288,0.007242666350470648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,16384,0.020053333706325956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,10240,0.00702311098575592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,51200,0.03864444295565287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,12288,0.017575111654069688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,10240,0.016837333639462788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,7168,0.006267555471923616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,8192,0.015192889504962497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,7168,0.014817777607176038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,6144,0.0064071110553211635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,6144,0.014126222994592456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,5120,0.005936000082227919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,5120,0.013515555196338229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,4096,0.0052800000541739995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,8192,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,3584,0.005896889087226655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,4096,0.012857777376969656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,3072,0.0052968888647026485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,3584,0.013028444515334235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,2560,0.005984888722499211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,3072,0.011855111353927188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,2048,0.005485333502292633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,2560,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,2048,0.011129777464601727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,1536,0.004307555655638377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,1024,0.003585777762863371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,1536,0.01127822200457255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,768,0.0035342222286595237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,1024,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,768,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,512,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,256,0.0028435554769304064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,512,0.010071111222108206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,128,0.0026266665922270883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,256,0.0107004443804423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,64,0.002526222210791376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,128,0.010409778190983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,128,32,0.002480000050531493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,64,0.010412444670995077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,128,32,0.009034666750166152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,65536,0.010337777435779572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,51200,0.009165333376990424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,65536,0.04587466518084208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,16384,0.006976000136799282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,51200,0.03842577669355605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,12288,0.006274666637182236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,16384,0.019746666153271992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,10240,0.005877333382765452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,12288,0.017904000149832833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,8192,0.006246222390068903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,10240,0.016534222496880423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,7168,0.005633777628342311
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,8192,0.015190222197108798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,7168,0.014793778459231058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,6144,0.005892444401979446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,5120,0.005601777798599667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,6144,0.014168888330459595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,4096,0.0052675555149714155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,3584,0.005600888695981767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,3584,0.012440889245933957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,3072,0.00536088893810908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,4096,0.013535111314720578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,2560,0.0059368887709246735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,3072,0.011776888536082374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,2048,0.005623999983072281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,2560,0.012774222426944308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,2048,0.011170667078759937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,1536,0.004552888787455029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,1024,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,1536,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,768,0.003576888806290097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,5120,0.013510222236315409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,512,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,1024,0.011847111086050669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,768,0.011110222174061669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,256,0.002975111206372579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,128,0.002798222212327851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,512,0.01071200023094813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,256,0.010436444646782346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,64,0.002524444419476721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,64,32,0.0025253333151340485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,64,0.009738666315873465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,65536,0.009076444639099969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,128,0.01144266708029641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,64,32,0.010023111270533668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,51200,0.008300444318188561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,16384,0.006623999940024481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,65536,0.045682665374543935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,12288,0.005906666732496685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,51200,0.03850222296184964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,16384,0.020263999700546265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,10240,0.006245333287451003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,12288,0.017619555195172627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,8192,0.005605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,7168,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,10240,0.016866667403115165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,8192,0.015166223049163818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,6144,0.005927111125654644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,5120,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,7168,0.014831999937693277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,6144,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,4096,0.0052933332820733385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,3584,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,5120,0.014123555686738757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,4096,0.012732444538010491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,3072,0.00564444437623024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,3584,0.012774222426944308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,2560,0.00598933340774642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,3072,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,2048,0.005286222116814719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,2560,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,1536,0.004435555388530095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,2048,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,1024,0.0036275556517971885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,1536,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,1024,0.01074222226937612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,768,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,512,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,768,0.011148444480366178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,256,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,512,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,128,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,256,0.010459555519951714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,64,0.002502222235004107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,128,0.011477332976129321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,64,0.009367999931176504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,6,32,32,0.0027644443843099806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,6,32,32,0.00904622260067198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,51200,0.4917910893758138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,51200,0.9644461737738715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,16384,0.3397200107574463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,12288,0.2457911173502604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,16384,0.24457777871025932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,12288,0.19368533293406168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,10240,0.20711733235253227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,8192,0.16867378022935656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,10240,0.15816799799601236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,7168,0.1499048868815104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,8192,0.1349440018335978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,6144,0.13106577926211888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,7168,0.10912266704771255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,5120,0.09726222356160481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,6144,0.10215377807617188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,4096,0.07724977864159478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,5120,0.08346133099661933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,3584,0.06862933105892606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,4096,0.07069244649675158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,3072,0.05946399768193563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,3584,0.06855022244983248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,2560,0.05074577861362033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,3072,0.05647199683719211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,2560,0.043599999613232084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,2048,0.041217777464124895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,1536,0.03191288974550035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,2048,0.04281333420011732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,1024,0.023063111636373732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,1536,0.03412177827623155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,768,0.018301332990328472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,1024,0.027161777019500732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,512,0.013386666774749756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,768,0.02323644525474972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,512,0.018586667047606576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,256,0.015240000353919135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,128,0.00600622221827507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,128,0.013549333645237817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,64,0.005666666560702854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,64,0.014179555906189812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,32,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,65536,32,0.012825777961148156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,65536,256,0.008947555389669206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,65536,0.5194986661275228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,65536,1.077789306640625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,51200,0.8464639981587728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,51200,0.3980115519629584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,16384,0.2737342251671685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,16384,0.15305333667331272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,12288,0.1996746725506253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,12288,0.15702577431996664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,10240,0.1699075566397773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,10240,0.12721333238813612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,8192,0.13975289132859972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,8192,0.09903644190894233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,7168,0.12211110856797959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,6144,0.10728800296783447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,7168,0.08867822090784709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,5120,0.09203377697202896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,6144,0.0842755569352044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,5120,0.06932710938983493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,4096,0.07624089055591159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,3584,0.06091022491455078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,4096,0.05952622493108114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,3072,0.05333955420388115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,3072,0.04861244559288025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,2560,0.045271111859215625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,3584,0.05039199855592516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,2048,0.03677599959903293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,2560,0.03720444440841675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,1536,0.02923822071817186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,1024,0.020426667398876615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,1536,0.029275556405385334
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,768,0.016523554921150208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,2048,0.036462220880720354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,512,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,1024,0.02327022287580702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,768,0.021703110800849065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,256,0.007391110890441471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,128,0.003975999852021535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,256,0.01349955548842748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,512,0.018089777893490262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,64,0.0032115555885765287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,128,0.012423111332787408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,51200,32,0.0034551111360390982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,32,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,51200,64,0.014294221997261047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,65536,0.32970044347974986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,65536,0.18060355716281465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,51200,0.25761511590745717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,16384,0.09673244423336452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,51200,0.13847466309865317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,12288,0.0687848859363132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,16384,0.07243377632564969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,10240,0.059353775448269315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,12288,0.0649155577023824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,10240,0.05228266782230801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,8192,0.049766222635904946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,7168,0.03693422343995836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,8192,0.04457688993877835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,6144,0.03237866693072849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,6144,0.035205334424972534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,7168,0.039988444911109075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,5120,0.027627555860413447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,4096,0.022711111439598933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,4096,0.026942221654786006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,3584,0.020255999432669747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,5120,0.031705776850382485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,3072,0.018111111389266122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,3584,0.026543110609054565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,2560,0.015533333023389181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,2560,0.019681778219011094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,3072,0.024332443873087566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,2048,0.013062222136391534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,1536,0.010784889260927836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,1536,0.018209778600268893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,1024,0.00812355594502555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,1024,0.015208888385030957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,768,0.0069191112286514705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,768,0.014217777384652032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,512,0.004616888860861461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,512,0.012635555532243518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,256,0.0035422220826148987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,256,0.011376000112957425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,2048,0.020290666156344943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,128,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,64,0.0028168888141711554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,64,0.010496888723638324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,128,0.012000888586044312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,16384,32,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,65536,0.24201599756876627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,16384,32,0.011454222102959951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,65536,0.15101510948605007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,51200,0.1886533366309272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,16384,0.07391822338104248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,51200,0.11924710538652207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,12288,0.05108799868159824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,16384,0.06772710879643758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,12288,0.054216000768873424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,10240,0.04444444510671827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,10240,0.04643199841181437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,8192,0.03717777795261807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,7168,0.03298399845759074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,8192,0.03859733210669623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,6144,0.028411553965674505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,7168,0.03506222367286682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,6144,0.030621333254708186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,5120,0.02548888822396596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,4096,0.020781333247820537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,5120,0.026919111609458923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,3584,0.019311111834314134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,4096,0.02568888829814063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,3072,0.017224889662530687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,3584,0.023815110325813293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,2560,0.015020444989204407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,3072,0.02241155505180359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,2048,0.01075466639465756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,2560,0.018603555030292936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,1536,0.00867733359336853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,2048,0.018957333432303537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,1536,0.016678222351604037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,1024,0.00700711127784517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,768,0.006057777752478917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,1024,0.013915555344687568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,512,0.0038453332251972626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,768,0.014449778530332776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,256,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,512,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,128,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,128,0.010781333678298526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,256,0.011863110793961419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,64,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,12288,32,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,32,0.010049777726332346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,12288,64,0.011707555916574268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,65536,0.20152000586191812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,65536,0.1477875577078925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,51200,0.1611164410909017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,16384,0.06218666500515408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,12288,0.04321600000063578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,51200,0.11530310577816433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,16384,0.05485777722464668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,10240,0.03758488761054145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,12288,0.045817779170142285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,8192,0.030991110536787245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,10240,0.03783644570244683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,7168,0.028339554866154987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,8192,0.03276177909639146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,6144,0.025617778301239014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,7168,0.030390222867329914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,5120,0.022050667140218947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,6144,0.02761777738730113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,5120,0.02364088926050398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,4096,0.018236445056067575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,3584,0.017256889078352187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,4096,0.021324444148275588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,3072,0.015376889043384127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,3584,0.020788444413079154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,2560,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,3072,0.01941688855489095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,2048,0.010703110860453712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,2560,0.01826755536927117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,1536,0.00869511150651508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,2048,0.016586666305859882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,1024,0.0069431112044387394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,1536,0.015544000599119397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,768,0.00544799988468488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,1024,0.013210666676362356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,512,0.004554666578769684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,768,0.013763555222087435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,256,0.0031591111587153543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,512,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,128,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,256,0.01182044463025199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,64,0.0028782222006056043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,128,0.011024889018800525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,10240,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,64,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,10240,32,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,65536,0.16628621684180364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,65536,0.09832355711195205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,51200,0.13043910927242702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,16384,0.04890666736496819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,51200,0.07679022020763822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,12288,0.03259999884499444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,16384,0.044778668218188815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,10240,0.028201778729756672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,12288,0.03636088967323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,8192,0.023139556248982746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,10240,0.031183110343085393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,7168,0.020655999581019085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,8192,0.026885333988401625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,6144,0.018339556124475267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,7168,0.02476088868247138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,5120,0.015999111864301894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,6144,0.023933332827356126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,4096,0.012894222305880653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,5120,0.020626667473051283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,3584,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,4096,0.018647111124462552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,3072,0.010749333434634738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,3584,0.018241778016090393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,2560,0.009660444325870937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,3072,0.017235555582576327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,2048,0.008447999755541483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,2560,0.016041777200169034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,1536,0.007173333730962541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,2048,0.014904888139830695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,1024,0.005250666704442766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,1536,0.014002665877342224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,768,0.003944000022278892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,1024,0.012797333300113678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,512,0.0038080000215106537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,768,0.013088888592190213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,256,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,512,0.011262222296661802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,128,0.0029155556112527847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,256,0.011401777466138205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,64,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,128,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,8192,32,0.0028942223224374983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,64,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,8192,32,0.01035644445154402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,65536,0.1440968910853068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,51200,0.1153706709543864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,65536,0.08894489208857219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,16384,0.04171644316779243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,51200,0.07049510876337688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,16384,0.04388711187574598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,12288,0.031102223528756037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,10240,0.02735200027624766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,12288,0.03606577714284261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,8192,0.022865777214368183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,10240,0.02981333269013299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,7168,0.020817778176731534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,8192,0.026471111509535048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,6144,0.016758221719000075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,5120,0.014896000425020853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,7168,0.02457511093881395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,4096,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,6144,0.023894222246276006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,5120,0.02008800043000115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,3584,0.011111111276679568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,4096,0.018599111172888014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,3072,0.010035555395815108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,3072,0.01736000014675988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,2560,0.009107555780145857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,2560,0.015576889117558798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,2048,0.007916444705592262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,2048,0.014456889695591398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,1536,0.006790222393141852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,3584,0.017729777428838942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,1024,0.004553777890072929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,1536,0.013461333182122974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,768,0.003587555554178026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,1024,0.012491555677519904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,768,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,512,0.0034728890491856467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,256,0.003162666741344664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,512,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,128,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,256,0.010743111371994019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,64,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,128,0.010812444819344414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,7168,32,0.002567111204067866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,64,0.010799110763602786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,7168,32,0.010144888526863521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,65536,0.12408888339996338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,51200,0.10045244296391805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,65536,0.08084622356626722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,16384,0.03973244296179877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,51200,0.06407911247677274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,12288,0.029134223858515423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,16384,0.041548444165123835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,10240,0.02481866710715824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,12288,0.033412443266974554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,8192,0.020690666304694284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,10240,0.026958222190539043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,7168,0.018583999739752877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,8192,0.026178666286998328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,6144,0.016857778032620747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,7168,0.023992889457278784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,5120,0.01534755527973175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,6144,0.023386667172114056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,5120,0.018910222583346896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,4096,0.011373333632946014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,3584,0.010492444038391113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,4096,0.017861333158281114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,3072,0.009501333038012186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,3584,0.017251556118329365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,2560,0.008390222158696916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,3072,0.016887111796273124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,2048,0.007662222617202335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,2560,0.014854222536087036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,1536,0.006326222171386083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,2048,0.014111111561457316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,1024,0.0042444442709287005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,1536,0.013165333204799227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,768,0.0035439998739295532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,1024,0.012441777520709567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,512,0.0032222221295038858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,768,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,256,0.003192000091075897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,512,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,128,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,256,0.010760000182522668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,128,0.011071110765139261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,64,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,6144,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,64,0.01038755559258991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,6144,32,0.010082667072614035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,65536,0.10480088657803005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,51200,0.08297955327563815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,65536,0.07322666380140516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,16384,0.034103999535242714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,51200,0.05890578031539917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,12288,0.025411556164423626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,16384,0.03786311215824551
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,10240,0.021549334128697712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,12288,0.030866665972603693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,8192,0.01899377836121453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,10240,0.025378665990299646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,7168,0.016750221451123554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,8192,0.024822221861945257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,6144,0.015075556106037565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,7168,0.021965333157115515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,5120,0.013392888837390475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,6144,0.02202133337656657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,4096,0.010439111126793755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,5120,0.018271999226676095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,3584,0.009738666315873465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,4096,0.01756533318095737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,3072,0.008671999805503422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,3584,0.016362667083740234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,2560,0.007972444097201029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,3072,0.015519110692871941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,2048,0.0069413334131240845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,2560,0.01385155568520228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,2048,0.013415111435784234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,1536,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,1536,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,1024,0.0038782221575578055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,1024,0.012143111063374413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,768,0.003510222252872255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,768,0.012094222009181976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,512,0.0031662223239739737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,512,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,256,0.011062222222487131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,128,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,128,0.011063999599880643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,64,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,64,0.01034400032626258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,5120,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,65536,0.08531910843319362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,5120,32,0.009689778089523315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,51200,0.06770044565200806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,65536,0.06614933411280315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,16384,0.027467555469936792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,51200,0.05381955703099569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,12288,0.020942222740915086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,16384,0.03120533294147915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,10240,0.01827022267712487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,12288,0.02815377712249756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,8192,0.0151235560576121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,10240,0.025109334124459162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,7168,0.014436443646748861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,8192,0.022967999180157978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,6144,0.013136000268989138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,7168,0.020554666717847187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,5120,0.011929777761300405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,6144,0.020263110597928364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,4096,0.010755555497275459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,5120,0.017070222232076857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,3584,0.010602666272057427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,4096,0.01648355523745219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,3072,0.009656888743241629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,3584,0.01570577753914727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,2560,0.008649778034951951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,3072,0.014510222607188754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,2048,0.007431999676757389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,2560,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,1536,0.005949333310127258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,2048,0.012732444538010491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,1024,0.004875555634498596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,1536,0.012788444757461548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,768,0.004211555752489301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,1024,0.012048888537618848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,512,0.004209777961174647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,768,0.011787555283970304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,256,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,512,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,128,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,256,0.01055111073785358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,64,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,128,0.011036444041464063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,4096,32,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,64,0.010010666317409938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,4096,32,0.00979466653532452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,65536,0.07498488823572795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,51200,0.06032533115810818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,65536,0.06352000104056464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,16384,0.023767110374238756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,51200,0.05205511053403219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,12288,0.018200000127156574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,16384,0.031926221317715116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,10240,0.01612622208065457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,12288,0.027354665928416785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,8192,0.01387733303838306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,10240,0.022628444764349196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,7168,0.013589333328935834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,8192,0.021306667062971327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,6144,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,7168,0.019928000039524503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,5120,0.011367110742463006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,6144,0.018584888842370775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,4096,0.010363555616802638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,5120,0.016490666402710807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,3584,0.010394666757848527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,4096,0.015267555912335714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,3072,0.00908088849650489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,3584,0.014494222071435718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,2560,0.008302222523424361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,3072,0.013920888304710388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,2048,0.00664444433318244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,2560,0.013422222601042854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,1536,0.0059075554211934405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,2048,0.012444444828563266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,1024,0.004651555584536659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,1536,0.012616888516479068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,768,0.004589333302444882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,1024,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,512,0.004512888689835866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,768,0.01203733351495531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,256,0.003154666680428717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,512,0.011047110789351992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,128,0.0028106667515304354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,64,0.002492444382773505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,128,0.010384889112578498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3584,32,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,64,0.01035199976629681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,256,0.01124533348613315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3584,32,0.009773333039548662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,65536,0.0648604432741801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,51200,0.05021333363321093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,65536,0.06061866548326281
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,16384,0.020980444219377305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,16384,0.02962755494647556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,51200,0.05052977800369263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,12288,0.016300444801648457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,10240,0.015266666809717814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,12288,0.023731556203630235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,8192,0.013105777402718862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,8192,0.02084000077512529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,7168,0.011968889170222811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,7168,0.01909955508179135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,6144,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,6144,0.018637332651350234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,5120,0.010341333018408881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,10240,0.021379555265108745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,4096,0.009722666607962714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,5120,0.015838222371207345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,3584,0.009029332962301042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,4096,0.014556444353527494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,3072,0.008010666403505538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,3584,0.013809777796268463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,2560,0.0075128889746136135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,3072,0.013830222189426422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,2048,0.005932444499598608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,2560,0.013266666895813413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,2048,0.012455110748608908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,1536,0.006086222413513396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,1024,0.004608889006906086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,1536,0.012512889173295764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,768,0.004289777742491828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,1024,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,512,0.003663111064169142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,768,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,256,0.003552888830502828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,512,0.011783999701340994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,256,0.011033777561452655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,128,0.002862222285734283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,64,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,128,0.01073688848151101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,3072,32,0.0028328889360030494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,64,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,65536,0.05477777785725064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,3072,32,0.00980711148844825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,51200,0.045088890526029796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,65536,0.058706667688157826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,16384,0.01993688941001892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,51200,0.0485155549314287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,12288,0.01497244503762987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,16384,0.02815999918513828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,10240,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,12288,0.02366222275627984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,8192,0.01204533295498954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,8192,0.0195795562532213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,10240,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,7168,0.011228444675604502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,6144,0.010572444233629437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,7168,0.018606222338146634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,6144,0.017159110969967313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,5120,0.009508444203270806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,4096,0.009279111193286048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,5120,0.014554666148291694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,3584,0.008545777863926357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,4096,0.014179555906189812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,3072,0.007667555577225155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,3584,0.013838222457302941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,2560,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,3072,0.01384622189733717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,2048,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,2560,0.01311377767059538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,1536,0.005603555589914322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,2048,0.012803555362754397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,1536,0.012089778151777057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,1024,0.005079999979999331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,768,0.004201777693298128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,1024,0.011746666497654386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,512,0.0036640001667870414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,768,0.012258666257063547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,512,0.011711111499203576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,256,0.0036888888312710654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,128,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,256,0.011072888970375061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,64,0.002526222210791376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,128,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2560,32,0.0026382222357723448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,64,0.01036444471942054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,65536,0.04515555500984192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2560,32,0.010792888700962067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,51200,0.03726844324005975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,65536,0.0565262238184611
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,16384,0.015018666783968607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,51200,0.047162665261162647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,16384,0.0251528885629442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,12288,0.01274222218328052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,10240,0.011770666473441653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,12288,0.02139555580086178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,10240,0.019634667370054457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,8192,0.011238222320874533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,8192,0.018061333232455783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,7168,0.010426667001512317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,7168,0.01555288831392924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,6144,0.009946666657924652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,6144,0.015252444479200574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,5120,0.009115555220180089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,5120,0.01443733274936676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,4096,0.008455110920800103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,4096,0.013751111096805997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,3584,0.006702222343948152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,3584,0.013248888982666863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,3072,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,3072,0.013207111093733044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,2560,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,2560,0.013233777549531726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,2048,0.0064186664919058485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,2048,0.012799111505349478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,1536,0.005672889037264719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,1536,0.012658666405412885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,1024,0.004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,1024,0.012462221913867526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,768,0.004282666577233208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,768,0.012155555188655853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,512,0.003607999947335985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,512,0.01202400028705597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,256,0.0032826666202810076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,256,0.011132444772455426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,128,0.0029377777957253983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,128,0.011429333024554782
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,64,0.011143999795118967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,2048,32,0.0028213332924577924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,2048,32,0.010739555789364709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,65536,0.0358168880144755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,65536,0.05553244551022848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,51200,0.02896711230278015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,16384,0.013325333595275879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,51200,0.04496977726618449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,12288,0.011490667031870948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,16384,0.024187554915746052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,10240,0.014209777116775513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,12288,0.020656888683636982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,8192,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,10240,0.01829688913292355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,7168,0.011847111086050669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,8192,0.016920000314712524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,6144,0.008377778033415476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,7168,0.015336889359686108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,5120,0.007297777467303806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,6144,0.014929778046078153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,4096,0.007319110963079665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,5120,0.013998222019937305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,4096,0.013541333377361298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,3584,0.013275555438465543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,3072,0.006119110931952794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,3072,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,2560,0.005640000104904175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,2560,0.012811555630630918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,2048,0.005983110931184556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,3584,0.0069360000391801195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,1536,0.004993777722120285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,2048,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,1024,0.004574222283230888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,1536,0.011799111134476133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,1024,0.012005333271291522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,768,0.0038924444880750445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,512,0.0035502223504914176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,768,0.012149333126015134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,256,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,512,0.010818666881985135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,128,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,256,0.011113777756690979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,64,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,128,0.010751999914646149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1536,32,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,64,0.011117333339320289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,65536,0.0251555558707979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1536,32,0.010038221875826517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,51200,0.020812445216708712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,65536,0.05275288886494107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,16384,0.014492443866199918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,51200,0.042872889174355396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,12288,0.011560888753996955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,16384,0.022006221943431433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,10240,0.010909332997269101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,12288,0.018664888209766813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,8192,0.009549332989586724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,10240,0.017222222354676988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,7168,0.009030222064918941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,8192,0.015621332658661736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,6144,0.007540444533030192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,5120,0.006974221931563483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,7168,0.015207111835479736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,4096,0.006992888947327931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,6144,0.014518222875065274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,5120,0.01349866638580958
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,4096,0.013444444371594323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,3072,0.006280888699822956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,3584,0.013382222089502545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,3072,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,2560,0.005949333310127258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,2560,0.012132444315486483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,2048,0.0063288890653186375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,2048,0.011778666741318174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,1536,0.0052826665341854095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,3584,0.006617777877383762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,1024,0.004298666699065102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,1536,0.012107555237081317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,768,0.004273777620659934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,768,0.011504888534545898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,1024,0.011391999820868174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,512,0.0038399998512532976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,256,0.0033031110134389666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,256,0.010418666733635796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,512,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,128,0.002941333378354708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,64,0.0029226665695508323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,128,0.010816000401973724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,64,0.009789333575301701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,1024,32,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,65536,0.021355556117163763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,1024,32,0.00942488925324546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,51200,0.01763288842307197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,51200,0.04247022337383694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,65536,0.05137422349717882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,16384,0.01237511138121287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,12288,0.010408889088365767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,12288,0.01799199978510539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,10240,0.009039999710188972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,10240,0.016739555531077914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,8192,0.00870577742656072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,8192,0.015231110983424716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,7168,0.008340444829728868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,7168,0.014362666341993542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,6144,0.007290666302045186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,16384,0.021200888686709937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,6144,0.014073777529928418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,5120,0.006626666833957036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,4096,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,5120,0.013267555170589022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,3584,0.0069520001610120135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,4096,0.012846222354306115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,3584,0.013184888495339287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,3072,0.0064106666379504735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,2560,0.0063440000845326324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,3072,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,2048,0.004932444542646408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,2560,0.012655110822783576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,2048,0.011443555355072021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,1536,0.004943111290534337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,1024,0.004259555704063839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,1536,0.011805333197116852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,768,0.0035991109907627106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,1024,0.011128889189826118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,512,0.003580444388919406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,512,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,256,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,256,0.011136000355084738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,128,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,768,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,64,0.0028311111446883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,128,0.01071733319097095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,768,32,0.002507555608948072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,64,0.009420444567998251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,65536,0.023541332946883306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,768,32,0.009769777456919352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,51200,0.01819022165404426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,65536,0.050219555695851646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,16384,0.01092266705301073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,12288,0.008660444782839881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,51200,0.042267554336123996
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,16384,0.0203484445810318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,10240,0.007668444679843054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,8192,0.008009778128729926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,10240,0.016673778494199116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,7168,0.00775733341773351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,8192,0.015607110328144498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,6144,0.007361777954631382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,7168,0.015212444795502557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,5120,0.00701155596309238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,6144,0.014507555299335055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,4096,0.0070017774899800616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,5120,0.013750221994188098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,3584,0.007063111497296228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,12288,0.017610667480362784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,3072,0.006335999816656113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,4096,0.012840888566441007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,2560,0.005982222242487802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,3072,0.012176888684431711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,2048,0.006019555446174409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,2560,0.01258399999803967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,2048,0.011389333340856763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,1536,0.004278222305907143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,1024,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,1536,0.011736888852384357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,768,0.0035582222044467926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,1024,0.011136888629860349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,768,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,512,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,3584,0.01348977784315745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,256,0.0032195556494924757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,512,0.010769777827792697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,128,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,256,0.011419555379284753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,64,0.00286311118139161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,128,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,64,0.010381333529949188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,512,32,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,65536,0.015222221612930298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,512,32,0.009754666851626502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,51200,0.012883555557992725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,65536,0.04692799846331278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,16384,0.008336889247099558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,51200,0.039650665389166936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,12288,0.007991111112965478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,16384,0.02000977761215634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,10240,0.007319110963079665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,10240,0.01660355594423082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,8192,0.0069875551594628235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,8192,0.015563555889659457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,7168,0.006673777682913675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,7168,0.014919999572965833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,12288,0.018038221531444125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,6144,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,5120,0.005933333188295364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,6144,0.014527999692493014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,4096,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,5120,0.013863999810483722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,3584,0.006000888844331105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,4096,0.01348622226052814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,3072,0.006106666806671355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,3072,0.012795555922720166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,3584,0.012830221818553077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,2560,0.0063279999627007385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,2048,0.005938666562239329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,2560,0.012449777788586087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,1536,0.004365333252482944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,2048,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,1024,0.003950222084919612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,1536,0.011414222419261932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,768,0.003546666767862108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,1024,0.01147555559873581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,512,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,768,0.01144088887506061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,256,0.003097777772280905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,512,0.01146577795346578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,256,0.01050222251150343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,64,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,128,0.011140444212489657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,256,32,0.002628444383541743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,64,0.010401777923107147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,256,32,0.009740444521109263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,65536,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,51200,0.011923555698659686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,65536,0.04664355516433716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,16384,0.007707555260923173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,51200,0.038535111480289035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,12288,0.007128000259399414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,16384,0.02029244436158074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,10240,0.007027555671003129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,12288,0.01774311065673828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,8192,0.006596444381607904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,10240,0.016938666502634685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,7168,0.006666666517655055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,8192,0.015558222929636637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,6144,0.006245333287451003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,7168,0.01514400045077006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,5120,0.005958222266700532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,5120,0.013595555391576556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,6144,0.014489778214030795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,4096,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,3584,0.005969777703285217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,4096,0.013073777986897362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,3072,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,3584,0.01347733371787601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,2560,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,3072,0.012458666331238218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,2560,0.01278222186697854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,2048,0.005590222362014983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,1536,0.004257777912749184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,2048,0.011785777906576792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,1024,0.003927999900446998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,1536,0.011629333098729452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,768,0.003524444583389494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,1024,0.011728888584507836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,512,0.0032239999208185407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,512,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,768,0.011216888825098673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,256,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,128,0.00254488881263468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,128,0.011144888897736868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,256,0.010444444086816577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,64,0.0025635556214385563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,128,32,0.002540444541308615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,64,0.01036888857682546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,128,32,0.010083555347389644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,65536,0.01038044442733129
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,65536,0.0463066664006975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,51200,0.009711999860074785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,51200,0.03870044483078851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,16384,0.0069644442862934535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,16384,0.020302222834693063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,12288,0.0063573333124319715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,12288,0.017913776967260573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,10240,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,10240,0.016903110676341586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,8192,0.006638222270541721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,8192,0.015511110424995422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,7168,0.006285333385070165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,7168,0.015133332875039844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,6144,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,6144,0.014465777410401238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,5120,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,5120,0.013462222284740873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,4096,0.005622222191757626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,4096,0.013456888496875763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,3584,0.005935110979610019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,3584,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,3072,0.005565333283609814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,3072,0.013040888640615674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,2560,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,2560,0.01275288893116845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,2048,0.005576888720194499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,2048,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,1536,0.004296888907750447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,1536,0.011713777979214987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,1024,0.0038684445122877755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,1024,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,768,0.003532444437344869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,768,0.011710222396585675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,512,0.0032266666077905228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,256,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,256,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,128,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,512,0.011920889218648275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,128,0.011007111105653973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,64,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,64,32,0.0028151110228565005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,64,0.010009778042634329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,65536,0.009015999734401703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,64,32,0.010650667051474253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,51200,0.008291555775536431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,65536,0.04622311062282986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,16384,0.006700444552633498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,51200,0.038782222403420344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,12288,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,16384,0.020248888267411124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,10240,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,12288,0.01795733306143019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,8192,0.005912888795137405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,10240,0.016520000166363187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,7168,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,8192,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,6144,0.005893333504597346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,7168,0.014896000425020853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,5120,0.005373333477311664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,6144,0.014779556128713818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,4096,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,5120,0.013728000223636627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,3584,0.005613333235184352
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,4096,0.013415999710559845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,3072,0.005610666755172942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,3584,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,2560,0.0059279998143513995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,3072,0.013054221868515015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,2560,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,2048,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,1536,0.00424799985355801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,1536,0.01184622198343277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,1024,0.0038613333470291565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,1024,0.01146666705608368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,768,0.0035137778355015647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,768,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,512,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,2048,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,256,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,512,0.011410666836632622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,128,0.0028497777465316984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,256,0.011136888629860349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,64,0.0028328889360030494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,128,0.011039999624093374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,5,32,32,0.002505777817633417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,64,0.011022222538789114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,5,32,32,0.009686222506894005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,51200,0.4913333257039388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,51200,0.969721794128418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,16384,0.33278489112854004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,16384,0.25831733809577095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,12288,0.24521867434183756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,12288,0.19758221838209364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,16384,0.18503555986616346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,10240,0.2064302232530382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,12288,0.1440275510152181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,10240,0.16171022256215414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,8192,0.1678453286488851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,8192,0.14145599471198186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,10240,0.1253191100226508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,8192,0.10593155357572769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,7168,0.1491760015487671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,51200,0.5136986838446723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,7168,0.11189777321285671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,6144,0.13060000207689074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,6144,0.10671199692620172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,7168,0.09505689144134521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,5120,0.11074666844473945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,5120,0.0824728873040941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,6144,0.08564977513419257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,4096,0.08249688810772367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,4096,0.06688977612389459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,5120,0.07476355632146199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,3584,0.07199644380145602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,4096,0.06520889202753703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,3584,0.062184002664354115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,3072,0.0620959997177124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,3584,0.06055111355251736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,3072,0.052705778015984424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,2560,0.05266222357749939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,2560,0.05055555701255798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,3072,0.055586669180128306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,2048,0.043339557117886014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,2560,0.05046400096681383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,2048,0.03883644607332017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,2048,0.04583644535806444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,1536,0.03348088926739163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,1024,0.023397333092159692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,1536,0.034408890538745456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,1536,0.04035022192531162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,768,0.018422222799725003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,1024,0.022720888257026672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,1024,0.0366213321685791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,768,0.02276266614596049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,512,0.01369244439734353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,768,0.030744890371958416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,256,0.008824889030721452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,512,0.018567999203999836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,512,0.029393778906928167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,128,0.006284444282452266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,256,0.025751110580232408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,256,0.015232889188660515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,64,0.005649777750174205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,128,0.014493332968817817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,65536,128,0.022089777721299067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,64,0.014173333843549093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,65536,32,0.01351199961370892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,65536,0.5195875697665745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,65536,32,0.005555555638339784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,65536,1.0776471032036674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,51200,0.8471333185831705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,51200,0.39804532792833114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,16384,0.2735999955071343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,16384,0.15170221858554417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,16384,0.14942755964067247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,12288,0.19862755139668783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,12288,0.12605599562327066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,65536,0.5246257781982422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,12288,0.11811110708448623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,51200,0.40360532866583926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,10240,0.16937332683139375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,8192,0.13887643814086914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,10240,0.13765599992540148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,8192,0.11205955346425374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,10240,0.10378666718800862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,7168,0.12239200539059109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,8192,0.08798933029174805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,6144,0.10719466871685451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,7168,0.09210399786631267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,7168,0.07992444435755412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,6144,0.07648533582687378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,5120,0.09180621968375312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,6144,0.0725146664513482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,4096,0.07558311356438531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,5120,0.06953244739108615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,5120,0.0638622244199117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,4096,0.05703910854127672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,3584,0.06722755564583673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,4096,0.05661333269543118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,3072,0.058952887852986656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,3584,0.05282488796446058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,3584,0.05288355549176534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,3072,0.045518222782346934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,2560,0.04442755712403191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,2560,0.04362044400639004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,3072,0.04994044370121426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,2560,0.04593777656555176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,2048,0.03717066513167487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,1536,0.02939288814862569
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,2048,0.0340008901225196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,2048,0.04298311140802172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,1024,0.020371556282043457
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,1536,0.02940711047914293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,1024,0.02088800072669983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,768,0.01588088936275906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,1024,0.032495998673968844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,768,0.02033955521053738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,768,0.028766221470303003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,512,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,512,0.027462222509913977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,512,0.016456888781653512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,256,0.007299555672539606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,256,0.013528000149461957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,1536,0.037695109844207764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,128,0.004015111260943943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,128,0.01346933344999949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,128,0.02274844381544325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,64,0.003625777860482534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,64,0.014212444424629211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,51200,256,0.02397777802414364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,51200,32,0.0035546666218174826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,51200,32,0.01240799989965227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,65536,0.3276124530368381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,65536,0.17897955576578775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,51200,0.2570506731669108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,51200,0.13757244745890299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,16384,0.0950968861579895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,16384,0.07215199867884318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,65536,0.18774222003089058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,12288,0.06845066944758098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,16384,0.06725333134333293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,12288,0.06016977628072103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,51200,0.1498204469680786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,10240,0.05892178085115221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,12288,0.05571111043294271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,10240,0.05383022295104133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,8192,0.049456000328063965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,10240,0.05180622140566508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,8192,0.04349955585267809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,7168,0.03686666819784377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,8192,0.046239998605516225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,6144,0.03224800030390421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,7168,0.04007822275161743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,7168,0.04422755704985725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,6144,0.03536799881193373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,5120,0.0277173337009218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,6144,0.04032266802257962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,4096,0.022760000493791368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,5120,0.03851999839146932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,5120,0.03173244330618117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,4096,0.027040888865788777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,3584,0.020015999674797058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,4096,0.03464444478352865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,3584,0.026085333691702947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,3072,0.01776088939772712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,3584,0.032461331950293645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,2560,0.015239111251301236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,3072,0.023795555035273235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,3072,0.031946665710873075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,2048,0.01310400002532535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,2560,0.022221333450741235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,2560,0.02885599931081136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,1536,0.010194666683673859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,2048,0.027787556250890095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,2048,0.01959022217326694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,1536,0.017799999978807237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,1024,0.008329778081840938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,1536,0.026774222652117412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,1024,0.015544889701737298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,1024,0.02635555631584591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,768,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,512,0.0053191110491752625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,768,0.023181334137916565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,512,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,512,0.022687110635969374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,256,0.0032346666687064697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,256,0.020032889313167997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,256,0.011915555430783166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,128,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,16384,128,0.019293333093325298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,64,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,128,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,768,0.006631111105283101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,16384,32,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,64,0.01013422260681788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,16384,32,0.010081777969996134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,65536,0.24246401256985137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,65536,0.15269954999287924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,51200,0.1875928905275133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,51200,0.11913511488172744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,16384,0.0724124444855584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,16384,0.06930222113927205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,16384,0.05761866437064277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,65536,0.14988444911109075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,51200,0.1206595500310262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,12288,0.052271998590893216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,10240,0.04506577716933357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,12288,0.049842665592829384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,10240,0.04646044307284885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,8192,0.037928889195124306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,12288,0.04972088999218411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,10240,0.04492622282769945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,8192,0.03772977656788296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,7168,0.03327911098798116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,8192,0.04115022222201029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,6144,0.028905779123306274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,7168,0.034155554241604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,7168,0.0389066669676039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,5120,0.02537422213289473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,6144,0.0351164440313975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,6144,0.03400444322162204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,5120,0.028579556279712256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,4096,0.021520000365045335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,5120,0.034789333740870156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,4096,0.02442488902144962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,3584,0.019295111298561096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,4096,0.0322382218307919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,3584,0.023011555274327595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,3584,0.031211555004119873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,3072,0.017298666967286002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,3072,0.028966221544477675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,3072,0.02145244512293074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,2560,0.015386665860811869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,2560,0.02843022346496582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,2560,0.020794666475719877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,2048,0.01016977760526869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,2048,0.026787555880016748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,2048,0.01829688913292355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,1536,0.009269333548016017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,1536,0.026153778036435444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,1536,0.01725955473052131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,1024,0.006598222172922558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,1024,0.022633777724372015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,1024,0.013863110707865821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,768,0.0057653333577844836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,768,0.021770666042963665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,768,0.013786666923099093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,512,0.003924444317817688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,512,0.021340444684028625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,512,0.012805333567990197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,256,0.0032071111102898917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,256,0.021365332934591506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,256,0.011726222104496427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,128,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,12288,128,0.021029333273569744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,128,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,64,0.0028355556229750314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,12288,32,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,64,0.01143200033240848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,12288,32,0.010096000300513374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,65536,0.20232800642649332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,65536,0.14531732930077448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,51200,0.16120089424981013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,51200,0.1152693298127916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,65536,0.13320977158016628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,16384,0.06349066893259685
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,16384,0.05681777662701077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,51200,0.10677244265874226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,12288,0.04432533184687296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,16384,0.05460888809627957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,12288,0.04380711250834995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,10240,0.03873599900139703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,12288,0.046527110868030123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,10240,0.03961777687072754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,10240,0.04338755541377597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,8192,0.0314062237739563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,8192,0.03842755489879184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,8192,0.034155554241604276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,7168,0.028576887316173975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,7168,0.029912001556820337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,7168,0.036052445570627846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,6144,0.025417778227064345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,6144,0.03536533315976461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,6144,0.02628799941804674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,5120,0.024129778146743774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,5120,0.03236000074280633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,4096,0.018215111560291715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,4096,0.021223111285103693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,5120,0.02215911116864946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,4096,0.030037333567937214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,3584,0.017464000317785475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,3584,0.020522667302025687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,3072,0.01574222246805827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,3584,0.03047910994953579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,2560,0.01180622229973475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,3072,0.02823644545343187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,3072,0.01942044496536255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,2048,0.010017777482668558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,2560,0.018776888648668926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,2560,0.027886221806208294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,2048,0.01642577846844991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,1536,0.00870311094654931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,2048,0.02812533246146308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,1024,0.006551110910044775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,1536,0.01510933372709486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,1536,0.02571644385655721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,768,0.005548444473081165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,1024,0.022357333037588332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,1024,0.013412444127930535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,768,0.012131555212868584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,512,0.004840888910823398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,768,0.022104889154434204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,256,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,512,0.02238044473859999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,512,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,128,0.002899555489420891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,256,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,256,0.02259644369284312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,64,0.0028133332315418455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,128,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,64,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,10240,32,0.00282222218811512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,10240,32,0.01069777790043089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,65536,0.1616471078660753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,10240,128,0.01998222205373976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,65536,0.0986257791519165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,51200,0.12960267066955566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,51200,0.07713066869311862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,65536,0.12120266755421956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,16384,0.04978044496642219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,16384,0.04277866747644213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,51200,0.09922844171524048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,12288,0.03251555562019348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,16384,0.05067288875579834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,12288,0.035475555393430926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,10240,0.028192000256644353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,12288,0.04392533169852363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,10240,0.03263466556866964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,10240,0.04123199979464213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,8192,0.02309155629740821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,8192,0.02716000046994951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,7168,0.020682666036817763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,7168,0.025099555651346844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,7168,0.03531022204293145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,6144,0.018272888329293992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,6144,0.022239110536045496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,6144,0.03332711259524027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,5120,0.01621955633163452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,8192,0.037632889217800565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,5120,0.02090311050415039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,5120,0.03323910964859857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,4096,0.01346666696998808
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,4096,0.030020445585250854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,4096,0.01900888813866509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,3584,0.018964444597562153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,3584,0.030088888274298772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,3072,0.011247110863526663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,3072,0.01756444407833947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,3584,0.011922666596041786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,3072,0.029247111744350855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,2560,0.01706755492422316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,2560,0.028528889020284016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,2048,0.008732444710201686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,2048,0.01519555515713162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,2560,0.009855999714798397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,2048,0.027416000763575237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,1536,0.007110222346252865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,1536,0.02338933282428318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,1536,0.01384622189733717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,1024,0.0052426668504873914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,1024,0.02202844454182519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,1024,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,768,0.003928889003064898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,768,0.02209866709179348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,768,0.012479999827014076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,512,0.003236444460021125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,512,0.022090666823916968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,512,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,256,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,256,0.020675554871559143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,256,0.011376889215575324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,128,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,64,0.003043555551105075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,8192,128,0.0199955552816391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,8192,32,0.0029013332807355454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,128,0.011742221812407175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,64,0.010016889207892949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,65536,0.14224178261227077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,8192,32,0.009720000127951304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,65536,0.08855288558536106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,51200,0.11495378282335068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,51200,0.07007822063234118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,16384,0.042281776666641235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,16384,0.045647998650868736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,51200,0.0990826686223348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,65536,0.11893867121802436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,12288,0.032645334800084434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,16384,0.049442665444480054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,10240,0.028162668148676556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,12288,0.04422933194372389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,12288,0.03443911009364658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,8192,0.023806222611003455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,10240,0.03008355696996053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,10240,0.04055555661519369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,7168,0.02126844392882453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,8192,0.027200000153647527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,8192,0.035383111900753446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,6144,0.016984888248973422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,7168,0.02441155579355028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,7168,0.035445332527160645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,6144,0.022044445077578228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,5120,0.014935111006100973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,6144,0.033542222446865506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,4096,0.012539555629094442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,5120,0.01994044416480594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,5120,0.03255733185344272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,3584,0.01110311100880305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,4096,0.018580444984965853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,4096,0.02882844540807936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,3072,0.010050666828950247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,3584,0.02942755487230089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,3584,0.01866311166021559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,3072,0.016873778568373788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,3072,0.028904000918070476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,2560,0.00887288898229599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,2560,0.027518222729365032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,2560,0.016711999972661335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,2048,0.007688889073001013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,2048,0.02646933330429925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,2048,0.014485332700941296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,1536,0.0064533332155810455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,1536,0.023043556345833674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,1536,0.013813333378897773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,1024,0.004772444566090901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,1024,0.02240000002914005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,768,0.003605333467324575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,1024,0.0124195557501581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,768,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,768,0.022679110368092854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,512,0.003567111161020067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,512,0.021731555461883545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,512,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,256,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,256,0.021124444074100916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,256,0.011182222101423474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,128,0.002893333426780171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,7168,128,0.019650666250122916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,128,0.011529777612951068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,64,0.0029066666546795103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,7168,32,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,64,0.01107377807299296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,7168,32,0.009506666825877296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,65536,0.12378044923146565
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,65536,0.08099644713931613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,51200,0.10028533140818278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,51200,0.06413955820931329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,65536,0.11832088894314235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,16384,0.039432889885372586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,51200,0.09873777627944946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,16384,0.04222755630811056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,12288,0.02994489007525974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,16384,0.05005155669318306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,12288,0.03087644444571601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,12288,0.043112887276543506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,10240,0.024832889437675476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,10240,0.03028800090154012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,10240,0.0398248897658454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,8192,0.020632889535692003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,8192,0.027149332894219294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,8192,0.035721778869628906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,7168,0.018957333432303537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,7168,0.023200889428456623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,6144,0.016785777277416654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,7168,0.03502755694919162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,6144,0.021271111236678228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,6144,0.034212443563673235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,5120,0.015189333094490899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,5120,0.031039112144046362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,5120,0.020806221498383414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,4096,0.011174221833546957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,4096,0.029633777009116277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,3584,0.010232888989978367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,4096,0.018439999885029264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,3584,0.01756888959142897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,3072,0.00941244430012173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,3072,0.016341333587964375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,3072,0.02749866743882497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,2560,0.008650666309727563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,2560,0.015415110521846347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,2560,0.02720355490843455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,2048,0.007185777856243982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,3584,0.02975289026896159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,2048,0.014378666877746582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,1536,0.006686222222116258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,2048,0.026233777403831482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,1536,0.013402666482660504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,1536,0.024051555328898962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,1024,0.00444088876247406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,1024,0.022266666094462078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,768,0.0038088887102074097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,768,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,768,0.02205955485502879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,512,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,512,0.011471110913488599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,512,0.021714667479197185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,256,0.0032231110251612137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,1024,0.012402666939629449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,256,0.011023999916182624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,128,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,256,0.022053332792388067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,128,0.011029333704047732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,64,0.0028479999552170434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,6144,128,0.021724444296624925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,6144,32,0.0027724444452259275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,64,0.010438222024175854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,65536,0.10523377524481879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,6144,32,0.010438222024175854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,65536,0.07315911187065972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,51200,0.0848826633559333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,51200,0.058902223904927574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,65536,0.11796977784898545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,51200,0.09873066345850627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,16384,0.03471555643611484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,16384,0.037457777394188776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,16384,0.049366222487555615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,12288,0.026170666019121807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,12288,0.02864444586965773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,12288,0.042656001117494374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,10240,0.021965333157115515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,10240,0.028232001596026953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,10240,0.039367109537124634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,8192,0.01899377836121453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,8192,0.03676266802681817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,8192,0.024887111451890733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,7168,0.016912000046836007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,7168,0.03493600090344747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,7168,0.022089777721299067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,6144,0.01534222231970893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,6144,0.033598221010631986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,6144,0.02030666669209798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,5120,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,5120,0.03231644299295213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,5120,0.019143111175960965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,4096,0.010503110786279043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,4096,0.030053334103690252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,4096,0.017128888103697035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,3584,0.010103999740547603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,3584,0.029576000240114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,3584,0.016864889197879367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,3072,0.00868977771864997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,3072,0.015189333094490899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,2560,0.008007999923494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,2560,0.013854222165213691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,2560,0.02629511058330536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,2048,0.006685333533419504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,2048,0.012889777620633444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,3072,0.028198222319285076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,1536,0.005495999836259418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,2048,0.023071999351183575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,1536,0.013164444102181328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,1536,0.022375111778577168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,1024,0.0042053332759274375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,1024,0.011415111521879831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,1024,0.023004444109068975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,768,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,768,0.01221777747074763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,768,0.021408889028761122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,512,0.0032515554792351196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,512,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,512,0.022387555903858606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,256,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,256,0.020999110407299463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,256,0.011128889189826118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,128,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,128,0.011732444167137146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,5120,128,0.020523554748959012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,64,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,5120,32,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,64,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,5120,32,0.010318222145239512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,65536,0.08667466375562881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,65536,0.0664293302430047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,51200,0.07037333647410075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,65536,0.11678755283355713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,16384,0.02793688906563653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,51200,0.05397422115008036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,51200,0.09744533565309312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,12288,0.021958221991856892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,16384,0.03202222122086419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,12288,0.027151111099455092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,12288,0.04147555430730184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,10240,0.01850222216712104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,10240,0.028186665640936956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,10240,0.039065778255462646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,8192,0.01575999955336253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,8192,0.035517334938049316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,8192,0.02400355537732442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,7168,0.014112000664075216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,16384,0.04786844385994805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,7168,0.021680000755521987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,7168,0.03401333424780104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,6144,0.01315199997689989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,5120,0.012172444827026792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,6144,0.02057066725360023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,6144,0.032116442918777466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,5120,0.01791200041770935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,4096,0.01069866700304879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,5120,0.03220622075928582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,4096,0.016548444827397663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,4096,0.02882933285501268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,3584,0.011448000040319232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,3072,0.009630222287442949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,3584,0.014861333701345654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,3584,0.027497778336207073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,2560,0.008300444318188561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,3072,0.01533422205183241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,3072,0.02580533259444767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,2048,0.007333333293596904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,2560,0.013796444568369122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,2560,0.02548355526394314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,2048,0.012787555654843649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,1536,0.005929777605666055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,2048,0.023048889305856492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,1536,0.012400888734393649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,1024,0.00526577772365676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,1536,0.022649778260125056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,1024,0.011852444046073489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,1024,0.021356445219781663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,768,0.004583111239804162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,768,0.02067377832200792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,768,0.011119999819331698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,512,0.003913777983850903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,512,0.020284444093704224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,512,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,256,0.0031679999083280563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,256,0.019550222489568923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,256,0.010696888797812991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,128,0.002805333377586471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,4096,128,0.018615111708641052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,64,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,64,0.009699555734793345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,4096,32,0.002536000063021978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,32,0.008977777428097194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,65536,0.07660977707968818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,65536,0.06265244219038221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,4096,128,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,51200,0.06230488750669691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,65536,0.11758844057718913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,51200,0.051615112357669406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,16384,0.02492800023820665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,51200,0.09768978092405532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,16384,0.047729776965247266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,16384,0.0306346681382921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,12288,0.02680355476008521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,12288,0.04148799843258328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,10240,0.016417778200573392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,10240,0.02647999922434489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,10240,0.03809955716133118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,8192,0.014012444350454541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,12288,0.018928888771269057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,8192,0.034663110971450806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,8192,0.02358666724628872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,7168,0.013191111385822296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,7168,0.020602666669421725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,7168,0.03309955530696445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,6144,0.012352888782819113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,6144,0.018556444181336295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,6144,0.032111111614439226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,5120,0.011711999773979187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,5120,0.01792888840039571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,5120,0.030019554826948378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,4096,0.00980088859796524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,4096,0.01550044450494978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,3584,0.01016533292002148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,4096,0.02902844548225403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,3584,0.014455111490355598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,3584,0.027141332626342773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,3072,0.009008000294367472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,3072,0.02379911144574483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,3072,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,2560,0.013423110875818463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,2560,0.024013333850436743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,2048,0.007089777953094906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,2048,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,2048,0.023044443792766992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,1536,0.005592000153329637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,2560,0.008007999923494127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,1536,0.012456888953844706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,1536,0.021361778179804485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,1024,0.004611555486917496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,1024,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,1024,0.021320889393488567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,768,0.00425866660144594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,512,0.005476444545719359
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,768,0.02068533334467146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,768,0.01107911103301578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,256,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,512,0.011457777685589261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,256,0.01071999967098236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,128,0.0029084444459941653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,256,0.020776889390415616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,128,0.011062222222487131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,64,0.0028168888141711554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,128,0.019606222709019978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3584,32,0.0025635556214385563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,64,0.010043555663691627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,65536,0.0674239993095398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3584,32,0.009402666654851701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,65536,0.06066933605406019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,51200,0.04987733231650459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3584,512,0.020038222273190815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,65536,0.11815022097693549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,51200,0.049738665421803795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,16384,0.020741333564122517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,51200,0.09750577476289536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,16384,0.02869510981771681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,16384,0.04833777745564779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,12288,0.01724799970785777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,10240,0.01515111161602868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,12288,0.024643555283546448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,12288,0.04205333193143209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,10240,0.022915555371178523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,8192,0.013419555293189155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,10240,0.039056890540652804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,8192,0.020407110452651978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,7168,0.012170666621790992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,7168,0.019336000084877014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,7168,0.03308622373474969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,6144,0.011619555453459421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,6144,0.018171555466122098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,6144,0.031517333454555936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,5120,0.010015111002657149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,8192,0.03526399864090814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,5120,0.016864000095261466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,4096,0.009677333136399588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,5120,0.03084533413251241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,4096,0.01516355574131012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,3584,0.0092604441775216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,4096,0.027388445205158655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,3584,0.014203555054134794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,3584,0.02644266684850057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,3072,0.008153777983453538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,3072,0.013451555536852943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,3072,0.023686221904224817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,2560,0.007163555257850223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,2560,0.012747555143303342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,2560,0.02402666707833608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,2048,0.006319111006127463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,2048,0.022635555929607813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,1536,0.005917333480384614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,1536,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,1536,0.02128977742460039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,1024,0.00489688871635331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,1024,0.01146488885084788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,1024,0.020690666304694284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,768,0.004568888909286923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,2048,0.012410666379663678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,768,0.010775110787815519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,512,0.004179555508825514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,768,0.021957332889238994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,512,0.010384889112578498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,256,0.0034933334423436057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,512,0.021624000536070928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,256,0.010344888601038191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,256,0.02095199955834283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,128,0.0028906667398081887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,128,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,3072,128,0.01849422189924452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,64,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,3072,32,0.0025048889219760895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,64,0.009647111097971598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,3072,32,0.009393778112199571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,65536,0.05835555659400093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,65536,0.1188231176800198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,51200,0.04779555400212606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,51200,0.04811111092567444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,16384,0.020657777786254883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,65536,0.058245334360334605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,16384,0.02647466626432207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,12288,0.014946666028764514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,51200,0.09742044740253025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,16384,0.04708710975117154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,10240,0.013611555927329592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,12288,0.023017777336968318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,12288,0.04214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,10240,0.021542222963439092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,8192,0.011748444702890186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,10240,0.03908266623814901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,7168,0.01090400003724628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,8192,0.019528888993793063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,8192,0.036144001616372004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,6144,0.010247111320495605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,7168,0.018537777993414137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,7168,0.03293422195646498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,5120,0.00941066692272822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,6144,0.01815555493036906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,6144,0.03208177619510227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,4096,0.0090506664580769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,5120,0.015832889411184523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,5120,0.030863109562132094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,3584,0.008654221892356873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,4096,0.014516444669829475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,4096,0.025806221697065566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,3072,0.008036444584528605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,3584,0.013763555222087435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,3584,0.026398221651713055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,2560,0.006631999793979857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,3072,0.012794666820102267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,3072,0.02539644473128849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,2048,0.005955555372767978
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,2560,0.012431999875439538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,2560,0.02304800020323859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,1536,0.005581333405441708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,2048,0.023560888237423364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,2048,0.012728888955381183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,1536,0.01180266671710544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,1024,0.005029333134492238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,1536,0.022650667362742957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,768,0.0038862224254343244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,1024,0.011446221835083432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,1024,0.02272533377011617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,768,0.01109777804878023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,512,0.00360000009338061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,768,0.022625777456495497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,512,0.011058666639857821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,256,0.0031902222997612427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,512,0.021996445126003687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,256,0.010785777535703448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,256,0.021943999661339655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,128,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,128,0.011393778026103973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2560,128,0.019659555620617337
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,64,0.002512000087234709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2560,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,64,0.010015111002657149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,65536,0.047691557142469615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2560,32,0.010081777969996134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,65536,0.05588799715042114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,51200,0.03940800163480971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,65536,0.11736355887518989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,51200,0.04607111215591431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,16384,0.015257777439223396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,51200,0.09745955467224121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,16384,0.02401155564520094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,12288,0.012510221865442065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,16384,0.047912001609802246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,12288,0.021381333470344543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,10240,0.011570666399266986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,12288,0.04180533356136746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,10240,0.02075199948416816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,10240,0.03886399997605218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,8192,0.011164444188276926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,8192,0.018941332896550495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,8192,0.035210665729310774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,7168,0.010393777655230628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,7168,0.016615110966894362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,7168,0.03429866830507914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,6144,0.010017777482668558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,6144,0.031934221585591636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,6144,0.015107555521859063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,5120,0.008815999660227034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,5120,0.014378666877746582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,5120,0.02854577700297038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,4096,0.008616000413894653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,4096,0.013783999615245394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,4096,0.027431999643643696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,3584,0.007642666498819987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,3584,0.02646488944689433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,3584,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,3072,0.0063484443558586975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,3072,0.02442577812406752
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,3072,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,2560,0.00590133335855272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,2560,0.02439022229777442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,2048,0.0063164445261160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,2560,0.012773333324326409
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,2048,0.01239822225438224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,2048,0.024192000428835552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,1536,0.005285333428117964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,1536,0.022639999787012737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,1536,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,1024,0.004580444345871608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,1024,0.022346667117542688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,1024,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,768,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,768,0.02167644434505039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,512,0.003930666794379552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,512,0.011369778050316704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,512,0.02163822286658817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,256,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,256,0.011092444260915121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,256,0.020957332518365648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,128,0.0029395555870400537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,768,0.011730666789743634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,128,0.01111288865407308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,64,0.002806222273243798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,2048,128,0.021643555826610986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,2048,32,0.0031582222630580268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,64,0.010047111246320937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,65536,0.03586311141649882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,65536,0.05511200096872118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,2048,32,0.010423111418883005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,51200,0.028945777151319716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,65536,0.11770578225453694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,51200,0.044750223557154335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,16384,0.013391999734772576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,51200,0.09769955608579849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,16384,0.04829422301716275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,16384,0.023659555448426142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,12288,0.010820444259378644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,12288,0.020503110355801053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,10240,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,10240,0.018940443793932598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,10240,0.03704444567362467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,8192,0.0121742222044203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,8192,0.017266665895779926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,12288,0.041856888267729014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,7168,0.011535999675591787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,8192,0.03404711021317376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,7168,0.015463999576038785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,7168,0.03217688865131802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,6144,0.00872622181971868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,6144,0.01483022173245748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,6144,0.03092355529467265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,5120,0.007808888951937358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,5120,0.014167111780908374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,5120,0.029808001385794744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,4096,0.007376888559924231
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,4096,0.013438222308953604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,4096,0.027268444498380024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,3584,0.00665244460105896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,3584,0.013342222405804528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,3584,0.026779555612140234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,3072,0.006295111030340195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,3072,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,3072,0.025807110799683466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,2560,0.00563733321097162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,2560,0.012840000291665396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,2560,0.025052444802390203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,2048,0.005919999960396025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,2048,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,1536,0.004934222333961063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,2048,0.022692445251676772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,1536,0.01223288890388277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,1024,0.004231111043029361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,1536,0.022299556268586054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,1024,0.012069332930776807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,1024,0.02166755497455597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,768,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,768,0.020982222424613103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,512,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,512,0.010430222584141625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,768,0.0038542221817705366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,512,0.021003555920388963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,256,0.003206222214632564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,256,0.010424888796276517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,256,0.02176977859603034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,128,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1536,128,0.02032800018787384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,128,0.011491555306646558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,64,0.0029306666304667792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1536,32,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,64,0.010728889041476779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1536,32,0.009464888936943477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,65536,0.025490666429201763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,65536,0.052256888813442655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,65536,0.11751822630564372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,51200,0.020788444413079154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,51200,0.04307377669546339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,16384,0.014156444205178155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,16384,0.02201955517133077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,16384,0.04807644420199924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,12288,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,12288,0.019015999303923715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,51200,0.09619288974338108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,10240,0.010471999645233154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,12288,0.04014222158326043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,10240,0.03638133406639099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,8192,0.00976533359951443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,10240,0.016965332958433364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,8192,0.015820443630218506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,7168,0.009405333134863112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,8192,0.032772445016437106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,7168,0.015513777732849121
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,7168,0.03129155437151591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,6144,0.014461333552996317
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,6144,0.029567999972237483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,5120,0.006616000086069107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,5120,0.01404622197151184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,4096,0.006986666884687211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,5120,0.027552889453040227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,6144,0.007743111087216272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,3584,0.006671999891599019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,4096,0.01277955538696713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,3584,0.026036444637510512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,3584,0.013568888935777875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,3072,0.006228444476922353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,3072,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,2560,0.005608888963858287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,2560,0.012812444733248817
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,2560,0.025078222155570984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,2048,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,2048,0.012432888978057437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,3072,0.024414221445719402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,2048,0.024047111471494038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,1536,0.012095111111799875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,4096,0.025455999705526564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,1536,0.023729777998394434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,1024,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,1024,0.020672000116772123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,1024,0.012039999994966718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,768,0.004276444514592488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,768,0.02036088870631324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,768,0.011707555916574268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,512,0.003929777691761653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,1536,0.006237333433495627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,256,0.003216888962520493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,512,0.01995733380317688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,512,0.011377777490350934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,128,0.0029333333174387612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,256,0.019299555155966017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,256,0.010370666782061258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,128,0.010468444062603844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,64,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,1024,128,0.020297777321603563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,1024,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,64,0.00943288869327969
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,65536,0.02069155540731218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,1024,32,0.009431111315886179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,65536,0.05084800057941013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,51200,0.017654221918847825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,65536,0.1158942249086168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,51200,0.042287998729281955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,16384,0.012172444827026792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,51200,0.09580000241597493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,16384,0.021567111214001972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,12288,0.01072533345884747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,16384,0.04592622319857279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,12288,0.017743999759356182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,10240,0.00924800005224016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,12288,0.03973955578274197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,10240,0.016550223032633465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,10240,0.03571733170085483
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,8192,0.00903911143541336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,8192,0.031897776656680636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,8192,0.01549600064754486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,7168,0.008766222331258986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,7168,0.030212445391549006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,7168,0.01477688882086012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,6144,0.00721688899728987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,6144,0.028559999333487615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,6144,0.014334221680959066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,5120,0.006638222270541721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,5120,0.013788444300492605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,5120,0.02751733362674713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,4096,0.006968888971540663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,4096,0.01313244468635983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,3584,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,4096,0.02609866691960229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,3584,0.012471111284361945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,3584,0.02549244463443756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,3072,0.006288888967699475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,3072,0.023319111929999456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,3072,0.012437333663304647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,2560,0.006242666807439592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,2560,0.02273066673013899
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,2560,0.011754666765530905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,2048,0.005223111146026188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,2048,0.022104000051816303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,2048,0.011408888631396823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,1536,0.004946666873163647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,1536,0.011404444773991903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,1024,0.004217777815130022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,1024,0.010805333654085795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,1024,0.021371554997232225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,768,0.003543111185232798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,768,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,1536,0.02108977735042572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,768,0.020976000361972388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,512,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,512,0.019813333948453266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,512,0.01038577738735411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,256,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,128,0.002999111182159848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,256,0.01945244438118405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,256,0.010419555836253695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,768,128,0.018592889110247295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,768,32,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,128,0.0110880004035102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,64,0.010028444230556488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,65536,0.022035555707083807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,768,32,0.008654221892356873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,51200,0.018970666660202872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,65536,0.05067111055056254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,65536,0.11695199542575413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,51200,0.04181066817707486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,16384,0.01057866629627016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,51200,0.09573066896862453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,12288,0.00869066682126787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,16384,0.01999022232161628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,16384,0.04602577620082431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,10240,0.0074373334646224976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,12288,0.017608889275126986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,12288,0.03843466771973504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,8192,0.007985778152942657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,10240,0.016894222961531747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,10240,0.036079999473359846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,8192,0.015318221516079374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,7168,0.007368000017272101
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,8192,0.03308533297644721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,6144,0.0069679998689227635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,7168,0.014495111174053617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,7168,0.03128977616628011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,5120,0.006600889066855113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,6144,0.014534221755133735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,6144,0.028800888193978205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,4096,0.0069724445541699724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,5120,0.013814222481515674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,5120,0.02821777926550971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,4096,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,3584,0.006670222100284364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,4096,0.026064000195927087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,3584,0.012493333054913415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,3584,0.02552355494764116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,3072,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,3072,0.023430221610599097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,3072,0.012412444584899478
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,2560,0.005967999911970562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,2560,0.0229404436217414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,2560,0.011796444654464722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,2048,0.005963555640644497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,2048,0.022061333060264587
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,2048,0.011437333292431302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,1536,0.004567111117972268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,1536,0.021031111478805542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,1024,0.003908444609906939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,1024,0.011101333631409539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,1024,0.020706666840447318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,768,0.003593777616818746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,768,0.01037688884470198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,768,0.020280889338917203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,512,0.003489777859714296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,1536,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,512,0.01036444471942054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,256,0.003198222153716617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,256,0.010120888551076254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,256,0.01921777758333418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,128,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,128,0.010390222072601318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,128,0.01865333318710327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,512,512,0.021033777130974665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,64,0.009343999955389235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,512,32,0.0025395556456512874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,65536,0.015174221661355762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,512,32,0.008532444636027018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,65536,0.0480968885951572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,51200,0.013173333472675748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,65536,0.11452355649736191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,51200,0.03903288973702325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,51200,0.09514844417572021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,16384,0.00832711077398724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,16384,0.019967110620604623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,12288,0.007967111137178209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,16384,0.04491733180152046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,12288,0.01757600075668759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,12288,0.039065778255462646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,10240,0.007340444458855524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,10240,0.03532266616821289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,10240,0.016551110479566786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,8192,0.006686222222116258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,8192,0.031946665710873075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,7168,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,7168,0.014427555931939019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,7168,0.030184000730514526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,6144,0.006260444306664997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,6144,0.013832000394662222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,6144,0.02847555610868666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,5120,0.005988444305128521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,8192,0.015146666102939181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,5120,0.013495999905798169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,4096,0.006253333141406377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,5120,0.028184887435701158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,4096,0.012843555874294706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,3584,0.005934222290913264
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,4096,0.025777777036031086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,3584,0.012442666623327466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,3072,0.005689777847793367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,3584,0.025756445195939805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,3072,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,2560,0.0059573331640826324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,3072,0.024458666642506916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,2560,0.011504888534545898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,2048,0.005320000151793162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,2560,0.023725334140989516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,2048,0.011967110964987012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,1536,0.0041991112132867174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,2048,0.023063111636373732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,1536,0.011189333266682096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,1536,0.022039999564488728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,1024,0.003690666622585721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,1024,0.020630222227838304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,1024,0.01073333372672399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,768,0.003536000019974179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,768,0.020642666353119742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,512,0.003160888950030009
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,512,0.0103502223889033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,512,0.019966221517986722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,256,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,256,0.010053333308961656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,256,0.018908444378111098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,128,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,768,0.010696888797812991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,128,0.010375111467308467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,64,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,256,128,0.02028888960679372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,256,32,0.002492444382773505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,64,0.009311999711725447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,65536,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,256,32,0.009316444396972656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,65536,0.046126223272747464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,65536,0.1143946647644043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,51200,0.012072888513406118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,51200,0.03843377696143256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,51200,0.09378577603234185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,16384,0.007641777396202087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,16384,0.01995733380317688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,16384,0.04496799906094869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,12288,0.007326222128338284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,12288,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,12288,0.038712888956069946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,10240,0.0069253332912921906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,10240,0.016574222180578444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,10240,0.035657776726616755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,8192,0.006611555400821898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,8192,0.015432000160217285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,7168,0.0063493334584765965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,8192,0.03360177742110358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,7168,0.014893333117167154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,6144,0.005959110955397288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,6144,0.014140443669425117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,7168,0.03173244330618117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,5120,0.006331555545330048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,6144,0.029487109846538965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,5120,0.01420088940196567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,4096,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,4096,0.013552000125249227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,4096,0.02650044361750285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,3584,0.006272000157170826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,3584,0.013424889081054263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,5120,0.027673777606752183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,3072,0.005589333259397083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,3584,0.02608977754910787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,3072,0.01239822225438224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,2560,0.005636444522274866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,3072,0.025761778155962627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,2560,0.025079111258188885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,2048,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,2048,0.012087110843923358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,2048,0.024068444967269897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,1536,0.004305777864323722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,1536,0.012054222325483957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,2560,0.011763555308183035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,1024,0.0037493333220481873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,1536,0.023676445086797077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,1024,0.010760000182522668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,768,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,1024,0.022720000810093347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,768,0.010767999622556897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,768,0.02238666680124071
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,512,0.0029493332323100832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,512,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,512,0.0207368897067176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,256,0.0030737777964936364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,256,0.010141332944234213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,128,0.0028880000528362063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,256,0.021731555461883545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,128,0.010873777998818291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,64,0.0026862221873468826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,128,32,0.002974222310715251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,64,0.010033778018421596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,65536,0.0107360002067354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,128,32,0.0094355551732911
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,51200,0.009870222045315636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,65536,0.04589955674277412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,16384,0.0069671107663048645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,51200,0.03872888949182298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,12288,0.006367110957702001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,16384,0.01964977714750502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,10240,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8_block,4,128,128,0.02030755579471588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,8192,0.006497777584526274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,12288,0.017824888229370117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,7168,0.006017777654859755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,10240,0.01668711172209846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,6144,0.005594666633341048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,8192,0.015810666812790763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,5120,0.005970666805903117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,7168,0.014900444282425774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,4096,0.005296000176005893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,6144,0.014127110441525778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,3584,0.005940444353553984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,5120,0.014172444740931192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,3072,0.005912000106440649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,4096,0.01348888874053955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,2560,0.0058106668293476105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,3584,0.012759110993809171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,2048,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,3072,0.013082666529549493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,2560,0.01219377749496036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,1536,0.004261333495378494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,1024,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,2048,0.011439111497667102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,1536,0.011750222080283694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,768,0.0035084444615576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,512,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,1024,0.011801777614487542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,256,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,512,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,256,0.011032889286677042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,128,0.0027377777215507296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,128,0.011490667031870948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,64,0.0025742221623659134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,64,0.010792888700962067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,64,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,768,0.011063111325105032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,65536,0.008988444175985124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,64,32,0.010064889159467485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,51200,0.008711111214425828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,65536,0.046332445409562856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,16384,0.006606222026877933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,51200,0.03872977693875631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,16384,0.02032533288002014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,12288,0.005973333285914526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,10240,0.005899555567238066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,12288,0.017889777819315594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,8192,0.005099555684460534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,10240,0.016917333006858826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,7168,0.006057777752478917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,8192,0.015577778220176697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,6144,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,7168,0.01516266663869222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,5120,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,6144,0.014166222678290473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,4096,0.005283555636803309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,5120,0.014193778236707052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,3584,0.005631999837027655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,4096,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,3072,0.005574222240183089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,3584,0.013200889031092325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,2560,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,3072,0.012738666600651212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,2048,0.005585777676767773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,2560,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,2048,0.012120000190205045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,1536,0.004520888957712385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,1024,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,1536,0.011717333561844297
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,768,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,1024,0.011735999749766456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,512,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,768,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,256,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,512,0.011366222467687396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,128,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,256,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,64,0.002487999904486868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,128,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,4,32,32,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,64,0.010037333601050908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,4,32,32,0.010060444474220276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,51200,0.4909084637959798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,16384,0.33189066251118976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,16384,0.1837528944015503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,12288,0.24429334534539116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,51200,0.9617119895087348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,10240,0.20611466301812065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,12288,0.20185600386725533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,10240,0.16595199373033312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,8192,0.1668142212761773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,8192,0.1175617774327596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,7168,0.14828622341156006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,6144,0.12938843833075628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,7168,0.09940799739625718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,5120,0.1092640029059516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,6144,0.09485777881410386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,4096,0.09038577477137248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,5120,0.08049599991904365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,3584,0.07215466764238146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,4096,0.06537422206666735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,3072,0.06217066446940104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,2560,0.05237866772545708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,3072,0.054006222221586436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,2048,0.04234222239918179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,2560,0.049321777290768094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,1536,0.032807111740112305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,2048,0.04189244574970669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,1536,0.030814223819308813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,1024,0.025754666990704004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,768,0.018271999226676095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,768,0.02235288918018341
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,3584,0.06172711319393582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,512,0.013719999955760108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,256,0.008773333496517604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,512,0.019114666514926486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,128,0.006248000181383557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,256,0.015159999330838522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,64,0.00554044461912579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,128,0.013443555268976422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,32,0.005555555638339784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,64,0.012809777425395118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,65536,1024,0.02320266597800785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,65536,32,0.0127697777416971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,65536,0.5184746848212348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,51200,0.8436284595065646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,51200,0.398144006729126
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,16384,0.26948711607191295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,16384,0.1503360006544325
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,12288,0.19938577546013725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,12288,0.1578400002585517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,10240,0.16866221692827013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,65536,1.0764924155341256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,8192,0.1382719940609402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,10240,0.13008622328440347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,7168,0.12178044848971897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,8192,0.10887733432981704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,6144,0.1073217789332072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,7168,0.10070933236016168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,5120,0.09097333086861505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,6144,0.07782577806048922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,4096,0.0749537746111552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,5120,0.07436622513665093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,3584,0.0669413341416253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,4096,0.06271288792292277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,3072,0.05863644679387411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,3584,0.050400889582104154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,2560,0.04454400142033895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,2560,0.0425875551170773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,2048,0.037019554111692644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,2048,0.036128888527552284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,1536,0.02870133188035753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,1536,0.028208000792397395
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,3072,0.043327109681235425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,1024,0.020187555087937247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,768,0.015871110889646742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,1024,0.022631999519136217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,512,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,768,0.020788444413079154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,256,0.007455999652544658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,512,0.01699733402993944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,128,0.003934222377008862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,256,0.014835556348164877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,64,0.003499555504984326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,128,0.012133333418104382
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,51200,32,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,64,0.013120888835854001
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,51200,32,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,65536,0.32672977447509766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,65536,0.1792595518959893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,51200,0.2563306755489773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,16384,0.09561510880788167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,51200,0.13743644290500218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,12288,0.06826044453514947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,12288,0.05911733044518364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,16384,0.07316444317499797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,10240,0.05852444304360283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,8192,0.04934666554133097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,10240,0.055042667521370776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,7168,0.036706666151682533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,8192,0.04542311032613119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,6144,0.03238577644030253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,7168,0.044159111049440175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,6144,0.035511109564039446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,5120,0.02751733362674713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,4096,0.022679999470710754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,5120,0.03287555442916022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,3584,0.020632889535692003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,4096,0.026756443911128577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,3072,0.017881777551439073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,2560,0.01573066744539473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,3584,0.0262746661901474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,2048,0.013079110946920184
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,3072,0.023873777853118047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,1536,0.01072799993885888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,2048,0.019858666592174105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,1024,0.008229333493444655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,1536,0.01827911039193471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,2560,0.021984888447655573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,768,0.0069368887278768755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,512,0.004932444542646408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,1024,0.014447111222479077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,256,0.003516444315512975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,512,0.013145777914259167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,256,0.011166222393512726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,128,0.003232888877391815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,768,0.013399110900031196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,64,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,16384,32,0.0028844444702068963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,128,0.010195555786291758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,64,0.010853333605660332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,65536,0.23981867896185982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,51200,0.1879813273747762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,65536,0.15294933319091797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,16384,0.07299644417232938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,51200,0.11886400646633571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,12288,0.05229777759975857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,16384,0.06050044298171997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,12288,0.04983733428849114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,16384,32,0.01070666644308302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,8192,0.0373564428753323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,10240,0.04261689053641426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,7168,0.03335377905103896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,8192,0.0433573325475057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,6144,0.0284222231970893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,7168,0.035659554931852556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,5120,0.025454221500290766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,6144,0.0311155551009708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,4096,0.021223111285103693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,5120,0.02854577700297038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,4096,0.02509333358870612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,10240,0.044673777288860746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,3584,0.0192248887485928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,3072,0.016912000046836007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,3584,0.02401688860522376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,3072,0.0214355554845598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,2560,0.015533333023389181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,2048,0.010681777364677854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,2560,0.02076088885466258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,1536,0.00868977771864997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,2048,0.018592889110247295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,1024,0.007045333584149678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,1536,0.016806221670574613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,768,0.005965333431959152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,1024,0.013520888984203339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,768,0.013481777575280933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,512,0.0037093332244290244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,256,0.003244444520937072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,512,0.012747555143303342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,128,0.0029102222373088202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,256,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,64,0.0029226665695508323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,128,0.01146755533085929
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,12288,32,0.0028888889484935333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,64,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,12288,32,0.010436444646782346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,65536,0.20204355981614855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,65536,0.14709866046905518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,51200,0.16054488552941218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,16384,0.06205155452092489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,51200,0.11476533942752415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,12288,0.04391555653678047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,16384,0.05342310998174879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,10240,0.03828444414668613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,12288,0.04500710964202881
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,8192,0.031007111072540283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,10240,0.040716442796919085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,7168,0.028344889481862385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,8192,0.03384977910253737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,6144,0.025685333543353613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,7168,0.02881066666709052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,5120,0.022077333596017625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,6144,0.027535999814669292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,4096,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,5120,0.02384977704948849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,3584,0.017172444197866652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,4096,0.022096888886557683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,3072,0.015711999601787992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,3584,0.020807999703619216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,2560,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,3072,0.019655999210145738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,2048,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,2560,0.018184888694021437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,1536,0.008852444589138031
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,2048,0.01741333305835724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,1024,0.0068897778789202375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,1536,0.014256889621416727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,768,0.005256888767083486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,1024,0.013772444592581855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,512,0.004589333302444882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,768,0.011772444678677453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,256,0.0029404444826973807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,512,0.012400888734393649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,128,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,256,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,64,0.002612444468670421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,128,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,10240,32,0.0029048888633648553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,64,0.010484444598356882
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,65536,0.1630773279401991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,65536,0.09523644712236191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,51200,0.12905332777235243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,51200,0.07632355557547675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,16384,0.04866310954093933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,10240,32,0.010434666441546546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,12288,0.0327413347032335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,16384,0.04152889053026835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,10240,0.028352889749738906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,12288,0.03631644447644552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,8192,0.023040889037979975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,10240,0.03297422329584757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,7168,0.02077155477470822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,8192,0.027857777145173814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,6144,0.018477333916558158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,7168,0.024122666981485155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,5120,0.016229333149062265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,6144,0.023323555787404377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,4096,0.013247999880048962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,5120,0.02035999960369534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,3584,0.012278222375445895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,4096,0.01927288870016734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,3072,0.010741333166758219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,3584,0.018258665998776753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,2560,0.009911999934249455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,3072,0.017581333716710407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,2048,0.008547555241319869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,2560,0.016562667157914903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,1536,0.007285333342022366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,2048,0.015840889679061044
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,1024,0.005256888767083486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,1536,0.01313333296113544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,768,0.003920888735188378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,1024,0.01256533298227522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,768,0.011536888778209686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,512,0.003661333272854487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,256,0.0032746667663256326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,512,0.011423110961914062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,256,0.011626666618718041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,128,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,64,0.0026071110947264563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,128,0.010392000277837118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,8192,32,0.0026355555488003623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,64,0.01016533292002148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,65536,0.14363555113474527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,8192,32,0.00943999985853831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,51200,0.11471022499932183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,65536,0.08851644727918838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,16384,0.04176888863245646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,51200,0.0700755582915412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,12288,0.03286666671435038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,16384,0.04028266668319702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,12288,0.03565688927968343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,10240,0.03267555435498556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,8192,0.02368533280160692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,8192,0.025781333446502686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,7168,0.021328000558747187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,7168,0.02342044479317135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,6144,0.016632889707883198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,10240,0.02789777848455641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,6144,0.022327999273935955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,5120,0.014783999986118741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,4096,0.01219466659757826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,5120,0.02034133341577318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,3584,0.011334222224023608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,4096,0.018246221873495314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,3072,0.010124444133705562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,3584,0.018424888451894123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,2560,0.009146666361225976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,3072,0.0162124451663759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,2560,0.01629244453377194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,2048,0.007999111380841997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,1536,0.0068862222962909276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,2048,0.014759999182489185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,1024,0.00461422238085005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,1536,0.012844444149070315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,768,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,1024,0.012452444268597497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,512,0.003590222034189436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,768,0.01142133358452055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,256,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,512,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,128,0.0028862222615215513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,256,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,128,0.010490666660997601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,64,0.0028444443725877335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,7168,32,0.00281866660548581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,64,0.010096000300513374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,7168,32,0.009378666679064432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,65536,0.12429689036475287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,51200,0.09980533520380656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,65536,0.08027733034557767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,16384,0.038888888226615057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,51200,0.06394755840301514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,12288,0.02974933385848999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,16384,0.03786133395300971
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,10240,0.025053333905008104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,12288,0.0325857765144772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,8192,0.02074044446150462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,10240,0.030798223283555772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,7168,0.019208888212839764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,8192,0.02419288953145345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,6144,0.016946666770511203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,7168,0.022309333086013794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,5120,0.01531555586391025
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,6144,0.0214008887608846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,4096,0.011244444383515252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,5120,0.019621334142155118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,3584,0.010649777948856354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,4096,0.01828266680240631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,3072,0.009476444787449306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,3584,0.01759999990463257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,2560,0.008591111335489485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,3072,0.01613333324591319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,2048,0.007663110891977946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,2048,0.014125333891974555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,2560,0.015302222636010913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,1536,0.0064071110553211635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,1024,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,1536,0.012804444465372296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,1024,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,768,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,512,0.0032008888406885993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,768,0.012235555383894177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,256,0.0031893334041039148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,512,0.011160000330872007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,128,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,256,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,64,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,128,0.010608889162540436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,64,0.009866666462686326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,6144,32,0.002846222163902389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,6144,32,0.009378666679064432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,65536,0.1050835582945082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,65536,0.0729093352953593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,51200,0.08426400025685628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,51200,0.05845955345365736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,16384,0.033747557136747576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,16384,0.03408799899948968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,12288,0.025964443882306416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,12288,0.029867556360032823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,10240,0.021695110532972548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,8192,0.01891377733813392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,10240,0.02846933404604594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,7168,0.01698933376206292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,8192,0.024104000793562994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,6144,0.015391111373901367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,7168,0.021350221501456365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,5120,0.013712888790501488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,6144,0.020270221763186984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,4096,0.010492444038391113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,5120,0.018917333748605516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,3584,0.009698666632175446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,4096,0.017649778061442904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,3072,0.009008000294367472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,3584,0.01664977769056956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,2560,0.008033778104517195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,3072,0.014538667268223233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,2048,0.007090667055712805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,2560,0.014204444156752693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,1536,0.0052355556852287715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,2048,0.013472889032628802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,1024,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,1536,0.012474666866991254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,768,0.00351555562681622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,1024,0.01143022212717268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,512,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,768,0.011071999867757162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,512,0.011512888802422417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,256,0.0029377777957253983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,128,0.0031315556002987754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,256,0.010761777559916178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,64,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,128,0.010533333652549319
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,5120,32,0.0026435556097163092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,64,0.00944088896115621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,65536,0.08704621924294366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,5120,32,0.009402666654851701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,51200,0.06981066862742107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,65536,0.06565244330300225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,16384,0.028371555937661067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,51200,0.053449776437547475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,12288,0.021321778496106465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,16384,0.0344177782535553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,10240,0.018752889500723947
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,12288,0.029816001653671265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,10240,0.023717333873112995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,8192,0.015824000040690105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,7168,0.014454222387737699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,8192,0.022289777795473736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,6144,0.013391111459996967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,7168,0.02108977735042572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,5120,0.011918221910794577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,6144,0.018645332919226754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,4096,0.01073066641887029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,5120,0.018664000762833487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,3584,0.010454222559928894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,4096,0.016347555650605094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,3072,0.009700444837411245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,3584,0.01553244392077128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,2560,0.00832711077398724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,3072,0.01348355578051673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,2048,0.0075777777367168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,2560,0.01420088940196567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,1536,0.0059004442559348206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,2048,0.0129120002190272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,1024,0.004923555586073133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,1024,0.011076444553004371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,768,0.004265777766704559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,1536,0.012465777496496836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,512,0.004229333251714706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,768,0.01108266661564509
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,256,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,512,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,128,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,256,0.010403555300500657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,64,0.0029111111329661477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,128,0.010749333434634738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,4096,32,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,64,0.009333333207501305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,65536,0.07671555545594957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,4096,32,0.009367110828558603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,51200,0.061632891496022545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,65536,0.06311821937561035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,16384,0.024752888414594863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,51200,0.05128711130883959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,16384,0.03233066532346938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,12288,0.02586577832698822
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,10240,0.01665155589580536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,10240,0.023016000787417095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,8192,0.014521777629852295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,12288,0.018903111418088276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,8192,0.021985777550273474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,7168,0.013423110875818463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,7168,0.020058666666348774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,6144,0.012693333129088083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,6144,0.018955555227067735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,5120,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,5120,0.017848889032999676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,4096,0.010062221851613786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,4096,0.0158906661801868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,3584,0.009749333063761393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,3584,0.014926221635606555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,3072,0.009017777939637503
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,2560,0.00797155582242542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,3072,0.014110222458839417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,2560,0.013778666655222574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,2048,0.00703022215101454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,1536,0.005945777727497949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,2048,0.0131217779384719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,1024,0.004583999928500918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,1536,0.012739555703269111
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,768,0.004442666553788715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,1024,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,512,0.005532444351249271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,768,0.012134222520722283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,256,0.0031235555393828284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,512,0.011100444528791638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,128,0.002855111120475663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,256,0.011724444727102915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,64,0.003133333391613431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,128,0.010729778144094678
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3584,32,0.0028604444944196274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,64,0.01070577816830741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3584,32,0.009116444322797988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,65536,0.06692444615893893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,51200,0.05080355538262261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,65536,0.06008444229761759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,16384,0.02163644466135237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,51200,0.0495013329717848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,12288,0.01660355594423082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,16384,0.028798222541809082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,10240,0.015198222464985318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,12288,0.025835555460717943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,8192,0.013095999757448832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,10240,0.023949333363109167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,7168,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,8192,0.021596444977654353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,6144,0.011095110740926532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,7168,0.018928888771269057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,5120,0.010384889112578498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,6144,0.01758399936887953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,4096,0.009773333039548662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,5120,0.0162435554795795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,3584,0.009171555439631144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,4096,0.014509333504570855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,3072,0.008336000144481659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,3584,0.014159111513031853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,2560,0.006981333096822103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,3072,0.013548444542619916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,2048,0.006305777778228124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,2560,0.013424889081054263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,1536,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,2048,0.013139555851618448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,1024,0.004885333279768626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,1536,0.012136889000733694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,768,0.004280000097221798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,1024,0.01203644441233741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,512,0.003926222109132343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,768,0.010856000085671743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,256,0.0035235554807715942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,512,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,128,0.002834666727317704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,256,0.011167111496130625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,64,0.002570666579736604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,128,0.01111822244193819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,3072,32,0.0025395556456512874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,64,0.010016000105275048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,3072,32,0.010056888891590966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,65536,0.05721244547102186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,51200,0.047446221113204956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,65536,0.058202664057413735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,16384,0.02035466664367252
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,51200,0.04751999841796028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,12288,0.015246222416559855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,16384,0.02648977769745721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,10240,0.014308444327778287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,12288,0.02458400030930837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,8192,0.012005333271291522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,10240,0.021848888860808477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,7168,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,8192,0.019280888968043856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,6144,0.010066666536860997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,7168,0.01898755629857381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,5120,0.009707555174827576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,6144,0.016214221715927124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,4096,0.00906044410334693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,5120,0.01496355566713545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,3584,0.008392000363932716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,4096,0.014204444156752693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,3072,0.008020444048775567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,3584,0.013789333403110504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,3072,0.013515555196338229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,2560,0.0069919998447100324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,2560,0.013199110825856527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,2048,0.006296000132958095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,2048,0.012763555679056378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,1536,0.005595555735958948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,1024,0.004658666749795278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,1536,0.01237155579858356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,1024,0.011877333124478659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,768,0.004379555583000183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,512,0.004008000095685323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,768,0.011689778003427716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,256,0.0034888887570963967
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,512,0.011566222541862063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,128,0.0029200000895394217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,256,0.011729777687125735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,64,0.002826666666401757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,128,0.010411555568377176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2560,32,0.0028977776981062363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,64,0.010773333410422007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,65536,0.048122667604022555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2560,32,0.00906488878859414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,51200,0.03974577784538269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,65536,0.05626133415434095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,16384,0.015190222197108798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,51200,0.046558221181233726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,12288,0.012783110969596438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,16384,0.024102222588327196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,10240,0.011663999822404651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,10240,0.019934222102165222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,8192,0.011120888921949597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,8192,0.018622222873899672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,7168,0.010703999963071613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,7168,0.016512889001104567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,6144,0.010045333041085137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,6144,0.015025777949227227
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,12288,0.021783111823929682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,5120,0.008996444443861643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,4096,0.007771555748250749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,5120,0.014153778553009033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,3584,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,4096,0.013831111292044321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,3072,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,3584,0.013153778182135688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,2560,0.005572444448868434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,3072,0.01312444441848331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,2560,0.012476444244384766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,2048,0.006588444527652528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,1536,0.005624889085690181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,2048,0.012082666986518435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,1024,0.004559999952713649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,1536,0.012180444267061023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,768,0.0039022221333450745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,1024,0.011381333072980246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,512,0.0035573335157500375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,768,0.011709333293967776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,256,0.0035093331502543557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,512,0.011436444189813403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,128,0.0032053333189752367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,256,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,64,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,128,0.010392000277837118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,2048,32,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,64,0.010686222049925061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,65536,0.03562577896647983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,65536,0.054692443874147206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,51200,0.02881866693496704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,51200,0.04452355702718099
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,16384,0.013356444736321768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,16384,0.023402666052182514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,12288,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,2048,32,0.009002666506502364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,12288,0.020608888732062448
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,10240,0.013767110804716745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,8192,0.011977777712874942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,10240,0.018907555275493197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,7168,0.010817777779367236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,8192,0.015570667054918079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,6144,0.008746667040718926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,7168,0.015469332536061605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,6144,0.014615999327765571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,5120,0.007337777978844113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,4096,0.007552000383536021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,5120,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,3584,0.0069617778062820435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,4096,0.013833777772055732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,3072,0.006224000205596288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,3584,0.013109332985348172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,2560,0.005605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,3072,0.01311377767059538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,2048,0.005889777921968036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,2560,0.012748444245921241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,1536,0.004907555464241239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,2048,0.012439111040698158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,1024,0.004321777572234471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,1536,0.012063999970753988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,768,0.004218666503826777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,1024,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,512,0.003601777884695265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,768,0.011739555332395764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,512,0.011426666544543372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,256,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,256,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,128,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,128,0.010397333237859938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,64,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,64,0.010045333041085137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1536,32,0.00906311141120063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,65536,0.02535111043188307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,65536,0.051799999343024365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,51200,0.020512888828913372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,51200,0.04267999860975477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,16384,0.014304889572991265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,16384,0.021713778376579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,12288,0.011381333072980246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,12288,0.017864000466134813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,10240,0.011035555766688453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,10240,0.01683200067943997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,8192,0.010044444766309526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,8192,0.015471110741297403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,7168,0.008985777695973715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,7168,0.01514933341079288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1536,32,0.00305600009030766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,6144,0.007652444144090016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,5120,0.006626666833957036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,5120,0.014224888549910652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,4096,0.006612444503439798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,4096,0.013416888813177744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,3584,0.006634666687912411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,3584,0.013233777549531726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,3072,0.006243555496136348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,3072,0.01274222218328052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,2560,0.00566133318675889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,2560,0.013367999758985309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,6144,0.014072888427310519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,2048,0.006589333216349284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,1536,0.006268444574541516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,2048,0.012136889000733694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,1536,0.012261333564917246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,1024,0.004753777964247597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,768,0.004594666676388847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,1024,0.011643555429246692
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,512,0.003887999802827835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,768,0.010877333581447601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,256,0.0034719999465677473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,512,0.01163200040658315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,128,0.0031155554784668815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,256,0.010500444306267632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,64,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,128,0.01037688884470198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,1024,32,0.002842666581273079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,64,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,1024,32,0.009015111459626092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,65536,0.0210506667693456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,51200,0.01756000022093455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,65536,0.050958222813076444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,16384,0.01184088902340995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,51200,0.042140444119771324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,12288,0.010398222340477837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,16384,0.02094488839308421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,12288,0.01794311073091295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,10240,0.00943199959066179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,8192,0.008979555633332994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,10240,0.0162435554795795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,7168,0.008307555483447181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,8192,0.015490666031837463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,6144,0.0069395556218094295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,7168,0.014533332652515836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,6144,0.014505777094099255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,5120,0.014169777433077494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,4096,0.007262222468852997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,4096,0.013419555293189155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,3584,0.006635555376609166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,3584,0.013063111239009433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,3072,0.006590222318967183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,3072,0.01272266689274046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,2560,0.006667555620272954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,5120,0.006624889042642381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,2048,0.004960888789759742
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,2560,0.012715555727481842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,1536,0.004889777965015835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,2048,0.011596444580290051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,1024,0.004490666505363253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,1536,0.012129777835475074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,768,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,1024,0.01033688916100396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,512,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,768,0.011374221907721626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,256,0.0028577778074476454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,512,0.010750222537252637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,128,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,256,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,64,0.002716444432735443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,128,0.009675555759006077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,64,0.009972444838947719
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,768,32,0.0028560000161329904
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,65536,0.023378666904237535
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,768,32,0.008973333570692275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,51200,0.019194665882322524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,65536,0.05035377873314751
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,51200,0.041843556695514254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,16384,0.010713777608341642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,16384,0.019998222589492798
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,12288,0.0083137775460879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,12288,0.017516444126764934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,10240,0.007626666790909237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,10240,0.016213332613309223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,8192,0.008023111356629265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,8192,0.014808000789748298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,7168,0.0076453329788313965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,7168,0.014498665928840637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,6144,0.007256888680987888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,6144,0.013733333183659447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,5120,0.0069351109365622205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,5120,0.013482666677898832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,4096,0.007239110767841339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,4096,0.012451555165979596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,3584,0.006614222294754452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,3584,0.012438221938080259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,3072,0.006421333385838403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,2560,0.006259555617968242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,3072,0.012087999946541257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,2048,0.0058906666106647914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,2560,0.012060444388124677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,2048,0.0114702218108707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,1536,0.004777777940034866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,1024,0.0035751110149754416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,1536,0.011432888607184092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,768,0.0035955554081334006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,1024,0.010755555497275459
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,512,0.0034862222770849862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,768,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,256,0.003175111073586676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,512,0.010718222293588849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,256,0.010409778190983666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,128,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,64,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,128,0.009454222189055549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,512,32,0.002573333266708586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,64,0.009696000152164036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,65536,0.015185778339703878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,512,32,0.008299555215570662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,51200,0.01279022213485506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,65536,0.046576887369155884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,16384,0.008298666940795051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,51200,0.03875022133191427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,12288,0.00794488853878445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,16384,0.019671999745898776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,12288,0.01724177764521705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,10240,0.007311999797821045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,8192,0.0069857777820693115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,10240,0.01624000072479248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,7168,0.006573333508438534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,8192,0.015487111277050443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,6144,0.006362666686375936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,7168,0.01445777714252472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,5120,0.0058755555914507965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,6144,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,4096,0.005939555664857228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,5120,0.013366222381591797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,3584,0.005670222143332164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,4096,0.012967999610635968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,3584,0.012488889197508493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,3072,0.005704888867007361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,2560,0.005598222215970357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,3072,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,2048,0.005576888720194499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,2560,0.012117333710193634
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,1536,0.004408888932731417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,2048,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,1024,0.003703111161788305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,1536,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,768,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,1024,0.010382222632567087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,512,0.003162666741344664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,768,0.01070666644308302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,256,0.002879111096262932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,512,0.010741333166758219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,128,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,256,0.010813333094120026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,128,0.009375111096435124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,64,0.002514666567246119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,256,32,0.0025475554996066624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,64,0.00869777798652649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,65536,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,256,32,0.009010666774378883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,51200,0.01183111137813992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,65536,0.04572622312439812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,16384,0.007314666277832455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,51200,0.03819377885924445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,12288,0.007304888632562425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,16384,0.01971733404530419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,10240,0.006627555522653792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,12288,0.017551110850440133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,8192,0.006640000061856375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,10240,0.016142222616407607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,7168,0.006249777972698212
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,8192,0.015220445063379077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,6144,0.005939555664857228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,7168,0.014456000592973499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,6144,0.01346933344999949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,5120,0.005591111050711737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,5120,0.013187555803192986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,4096,0.005586666779385672
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,4096,0.012402666939629449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,3584,0.005738666488064661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,3072,0.0053084443012873335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,3072,0.012092444631788464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,2560,0.005930666708283954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,3584,0.013776000175211163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,2560,0.012115555504957834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,2048,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,1536,0.004263111286693149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,1536,0.011676444775528379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,1024,0.003752888904677497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,1024,0.010379555324713389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,768,0.0035742223262786865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,768,0.01075644459989336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,2048,0.005576888720194499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,512,0.003301333429084884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,256,0.002987555538614591
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,256,0.010392000277837118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,512,0.011483555866612328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,128,0.0028719999310043124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,64,0.0025519999778932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,128,0.010417777631017896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,128,32,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,32,0.009089777866999308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,65536,0.01110666659143236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,65536,0.04627111223008898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,51200,0.00978311151266098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,128,64,0.009000889129108852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,16384,0.006342222293217977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,51200,0.03873599900139703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,16384,0.01960266629854838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,12288,0.006308444258239534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,10240,0.00598933340774642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,12288,0.017292444904645283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,8192,0.006269333263238271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,10240,0.01685244507259793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,7168,0.005977777971161737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,8192,0.01494488947921329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,6144,0.005616888817813661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,7168,0.015229332778188916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,6144,0.013515555196338229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,5120,0.005632888939645555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,4096,0.005605333381228977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,5120,0.013159111142158508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,3584,0.005621333503060871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,4096,0.013447110851605734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,3584,0.012495111260149213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,3072,0.005547555370463266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,2560,0.005997333261701796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,3072,0.012081777883900536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,2048,0.005603555589914322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,2560,0.012479110724396177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,1536,0.004514666481150521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,1536,0.012144000165992312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,1024,0.0037715555065208008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,1024,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,768,0.0036559998989105225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,768,0.011821333732869891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,512,0.0032879999942249725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,512,0.011588444312413534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,256,0.0032293332947625052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,256,0.011525332927703857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,128,0.0029608888758553397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,128,0.01053066634469562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,2048,0.011384888655609555
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,64,0.0025822222232818604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,64,32,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,64,0.00905600024594201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,65536,0.0087013335691558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,64,32,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,51200,0.008373333348168267
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,65536,0.04539733462863498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,16384,0.006663110935025745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,51200,0.038796444733937584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,16384,0.019718221492237516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,12288,0.005946666830115848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,10240,0.005660444498062134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,12288,0.017274666163656447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,8192,0.005600000007285013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,10240,0.016889777448442247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,7168,0.005712888720962737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,8192,0.01478844384352366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,6144,0.005618666609128316
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,7168,0.015202666322390238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,6144,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,5120,0.005401777724424998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,4096,0.005320888840489917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,5120,0.013094222380055321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,3584,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,4096,0.01314133322901196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,3584,0.012441777520709567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,3072,0.011800000237094032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,2560,0.005655111124118169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,2560,0.01218577805492613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,2048,0.0053013331360287136
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,2048,0.012191111014948951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,3072,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,1536,0.004401777767472797
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,1024,0.0038515557017591265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,1024,0.011113777756690979
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,1536,0.011875555747085147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,768,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,512,0.00318755561278926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,768,0.011743110915025076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,512,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,256,0.0032755554550223877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,256,0.011086222198274402
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,64,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,128,0.010402667025725046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,64,0.009704888694816166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,3,32,32,0.0025555555605226094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,3,32,32,0.008996444443861643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,51200,0.5414569112989638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,51200,0.9601092868381076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,16384,0.3288888931274414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,12288,0.24280267291598848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,16384,0.18707111146714953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,10240,0.20414400100708008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,12288,0.1509902212354872
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,8192,0.1659031046761407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,10240,0.14897955788506403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,7168,0.1465244425667657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,8192,0.11644355456034343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,7168,0.10574666659037273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,6144,0.12884710894690618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,5120,0.10947910944620769
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,6144,0.09573066896862453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,5120,0.08296266529295179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,4096,0.06908533308241102
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,3584,0.07190310955047607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,3584,0.06410844458474053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,3072,0.06207199891408285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,3072,0.05355911122428047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,2560,0.05261511272854275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,4096,0.08927644623650445
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,2048,0.04272266560130649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,2560,0.04786933461825053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,1536,0.033168001307381526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,2048,0.0388080014122857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,1536,0.034703999757766724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,1024,0.02308355602953169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,768,0.01850044396188524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,1024,0.02590933276547326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,512,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,512,0.018921777606010437
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,768,0.02277155551645491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,256,0.00866844422287411
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,128,0.005972444597217772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,256,0.016213332613309223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,128,0.013198222551080914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,64,0.005589333259397083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,65536,32,0.005616000129116907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,64,0.01293244461218516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,65536,32,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,65536,0.5182551278008355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,65536,1.0812533696492512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,51200,0.8436746597290039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,51200,0.43209245469835067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,16384,0.271815988752577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,16384,0.19882844554053414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,12288,0.19790844122568765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,12288,0.12372355990939671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,10240,0.1682008902231852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,10240,0.10869599713219537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,8192,0.09410844246546428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,7168,0.1206631130642361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,7168,0.08655466636021932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,6144,0.10591999689737956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,8192,0.1381937795215183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,6144,0.07852088742785983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,5120,0.09036799934175278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,4096,0.07452444235483806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,5120,0.06822222471237183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,3584,0.06618577904171415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,4096,0.05786755349900988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,3584,0.054442667298846774
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,3072,0.0461404456032647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,2560,0.04428622126579285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,2560,0.04068355427847968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,2048,0.03622577918900384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,2048,0.03323377834426032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,1536,0.028167999453014795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,1536,0.029020445214377508
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,1024,0.020584889584117465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,3072,0.058435552650027804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,1024,0.022986667023764715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,768,0.015791111522250705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,512,0.011379555695586734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,768,0.020813332663642038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,256,0.007541333635648091
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,512,0.01725955473052131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,128,0.0038720000949170855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,256,0.014454222387737699
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,64,0.0035111109415690103
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,128,0.012396444049146442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,64,0.012321777641773224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,51200,32,0.0034808889031410217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,51200,32,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,65536,0.32751467492845326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,65536,0.17943289544847277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,51200,0.2040568855073717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,16384,0.09481955236858791
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,16384,0.07447466585371229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,12288,0.06802311208513048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,51200,0.2563982274797228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,10240,0.058634665277269155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,12288,0.06198310852050781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,8192,0.048510223627090454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,10240,0.05113155643145243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,7168,0.037159999211629234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,8192,0.04747910963164436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,6144,0.0329786671532525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,7168,0.04109777675734626
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,6144,0.03734666771358914
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,5120,0.027600889404614765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,4096,0.02276444435119629
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,5120,0.03231733375125461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,3584,0.02072533302836948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,4096,0.02665688925319248
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,3072,0.017847110827763874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,3584,0.026478222674793665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,2560,0.015820443630218506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,3072,0.023559111687872145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,2560,0.02236088944805993
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,2048,0.013017777767446307
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,1536,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,2048,0.019307555423842538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,1024,0.008285333712895712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,1536,0.017616889543003507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,768,0.007080000307824876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,1024,0.01587199999226464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,512,0.004951111144489712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,768,0.013528889252079858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,256,0.0031804444475306403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,512,0.01276000009642707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,128,0.0029031110720502008
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,256,0.011863999896579318
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,64,0.0028613333900769553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,128,0.010492444038391113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,16384,32,0.0029093333416514923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,64,0.010116444693671333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,16384,32,0.009761778016885122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,65536,0.2396311230129666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,51200,0.18726044230990938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,65536,0.15130578147040472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,16384,0.07205244567659166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,51200,0.17411911487579346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,12288,0.051944000853432544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,10240,0.044738666878806226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,16384,0.0691359970304701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,12288,0.05246577660242716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,8192,0.0378479990694258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,10240,0.04610400067435371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,7168,0.033300442828072443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,8192,0.04268177681499057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,6144,0.030224889516830444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,5120,0.026379555463790894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,6144,0.031770666440327965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,7168,0.03476266728507148
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,4096,0.021346666746669348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,3584,0.019331556227472093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,5120,0.027109333210521277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,4096,0.025252444876564875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,3072,0.017699556218253244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,3584,0.023709333605236475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,2560,0.01609599921438429
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,2048,0.010431110858917236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,2560,0.020258666740523446
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,3072,0.022416000564893086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,1536,0.008842666943868002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,1024,0.007108444141017065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,2048,0.01740711099571652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,768,0.0059164443777667145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,1536,0.016837333639462788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,1024,0.013380444712109037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,512,0.003648888733651903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,256,0.003289777785539627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,512,0.011438222395049201
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,768,0.012816000315878125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,64,0.0029146667155954572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,256,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,12288,32,0.002938666691382726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,128,0.010679999987284342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,64,0.0090524446633127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,12288,32,0.009932444327407414
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,65536,0.19986755318111846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,65536,0.14557777510748968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,51200,0.15911289056142172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,16384,0.06208178069856432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,12288,0.04344888859324985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,51200,0.1288613345887926
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,10240,0.037603557109832764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,12288,0.043663998444875084
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,10240,0.03854044609599643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,8192,0.03071555495262146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,8192,0.031694223483403526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,7168,0.028035554620954726
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,16384,0.05537777807977465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,7168,0.030172444052166406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,6144,0.025603555970721777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,5120,0.022330666581789654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,6144,0.026921777261628047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,4096,0.018385777870814007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,5120,0.02490666674243079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,3584,0.01791822248035007
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,4096,0.021311110920376245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,3072,0.01588622232278188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,3584,0.02144266664981842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,2560,0.012168000141779581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,3072,0.01955111159218682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,2048,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,2560,0.018696000178654987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,1536,0.00869688888390859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,2048,0.016069332758585613
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,1024,0.007059555914666917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,1536,0.01534222231970893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,768,0.005398222141795688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,1024,0.012568888564904531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,512,0.004937777916590373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,768,0.013426666458447775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,256,0.00319377767542998
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,512,0.011434666812419891
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,128,0.0029644444584846497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,256,0.011781333221329583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,64,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,128,0.010481778118345471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,10240,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,64,0.009724444813198512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,65536,0.16243733300103083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,10240,32,0.009659555223253038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,51200,0.12897333833906385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,65536,0.0985493328836229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,16384,0.0495742228296068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,51200,0.09541333383984035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,12288,0.033878223763571844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,16384,0.042528887589772545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,10240,0.028295109669367473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,12288,0.033950222863091364
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,8192,0.023990222149425085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,10240,0.030900445249345567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,7168,0.021009777983029682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,8192,0.025085333320829604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,6144,0.018909333480728995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,7168,0.024910221497217815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,5120,0.01616533266173469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,6144,0.023007111416922674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,4096,0.013757333159446716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,5120,0.020974222156736586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,3584,0.012226666841242047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,4096,0.018601778480741713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,3072,0.011132444772455426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,3584,0.018625777628686693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,2560,0.009725333087974125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,3072,0.017504000001483493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,2048,0.008648888932334052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,2560,0.016892444756295946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,1536,0.007284444239404466
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,2048,0.01419911119672987
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,1024,0.00536088893810908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,1536,0.013571555415789286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,768,0.0038524443904558816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,1024,0.012129777835475074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,512,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,768,0.01238666640387641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,256,0.0031431110368834603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,512,0.011167111496130625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,128,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,256,0.011395555403497485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,64,0.0029324444217814338
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,128,0.010392888552612728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,8192,32,0.0026942222482628296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,64,0.009362666971153682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,8192,32,0.009398221969604492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,65536,0.14294577969445124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,51200,0.11390488677554661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,65536,0.0893484420246548
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,16384,0.042882667647467725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,51200,0.09869333108266194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,12288,0.03231644299295213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,16384,0.04576355550024244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,10240,0.028031110763549805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,12288,0.03637155559327867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,8192,0.02369244396686554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,10240,0.029230223761664495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,7168,0.021316443880399067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,8192,0.026629333694775898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,6144,0.01680266691578759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,7168,0.024296000599861145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,5120,0.014801777071423002
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,6144,0.022651554809676275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,4096,0.012469333079126147
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,5120,0.020364445116784837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,3584,0.011382222175598145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,4096,0.018744889232847426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,3584,0.018560889694425795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,3072,0.01016977760526869
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,3072,0.017575111654069688
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,2560,0.009361777868535783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,2560,0.01644177734851837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,2048,0.008375110725561777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,1536,0.007059555914666917
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,2048,0.0140657772620519
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,1024,0.004359111189842224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,1536,0.013544888959990608
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,768,0.003944000022278892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,1024,0.012506666282812754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,768,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,512,0.0035786665976047516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,256,0.0032622222271230486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,512,0.011431111229790581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,128,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,256,0.011594666375054253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,64,0.0028657778683635923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,128,0.010103999740547603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,7168,32,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,64,0.009701333112186855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,7168,32,0.009420444567998251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,65536,0.12367199526892768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,51200,0.09954310788048638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,65536,0.08033511373731825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,16384,0.03894666830698649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,16384,0.043024000194337636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,51200,0.08679378032684326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,12288,0.02988622254795498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,10240,0.02512888941499922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,10240,0.02760800056987339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,8192,0.021057777934604224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,8192,0.024526221884621516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,7168,0.01936444474591149
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,7168,0.023392889234754775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,6144,0.01758311192194621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,6144,0.02172800070709652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,5120,0.015712888704405892
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,12288,0.03423022230466207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,4096,0.011377777490350934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,4096,0.018179555733998615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,3584,0.010431999961535135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,3584,0.017987555927700467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,3072,0.009418666362762451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,3072,0.01685422162214915
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,2560,0.00868088917599784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,5120,0.019292443990707397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,2560,0.015146666102939181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,2048,0.013832888669437833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,1536,0.0064035554726918536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,1536,0.013440888788965015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,1024,0.004271999829345279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,1024,0.012429333395428128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,768,0.0038408889538711975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,2048,0.007640000018808577
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,512,0.0035146665241983202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,768,0.01237777786122428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,256,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,512,0.011378666592968835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,128,0.0028640000770489373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,256,0.011415999796655444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,64,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,128,0.009851555857393477
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,6144,32,0.002551111082235972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,64,0.009689778089523315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,6144,32,0.009043555292818282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,65536,0.10452533430523342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,51200,0.08404444323645698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,65536,0.07213866710662842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,16384,0.033928000264697604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,51200,0.07740622096591525
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,16384,0.03840177920129564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,12288,0.026038222842746313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,10240,0.02217688825395372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,12288,0.02812266680929396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,8192,0.0192239996459749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,10240,0.02718488872051239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,7168,0.017210667332013447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,8192,0.023129777775870428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,6144,0.015642666154437594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,7168,0.021779555413458083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,5120,0.013796444568369122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,5120,0.018925334016482036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,4096,0.0107351111041175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,4096,0.017653332816229928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,3584,0.009713778065310584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,3584,0.01717866626050737
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,3072,0.00886399961180157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,6144,0.02072177827358246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,2560,0.008053333395057255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,2048,0.00740533322095871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,3072,0.015171556009186639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,2560,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,1536,0.0052826665341854095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,2048,0.01347644461525811
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,1024,0.003906666818592283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,768,0.003563555578390757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,1536,0.012763555679056378
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,1024,0.011511999699804517
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,512,0.0032213332338465583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,768,0.012074666718641916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,256,0.002921777880854077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,512,0.010789333118332757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,128,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,256,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,64,0.002667555585503578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,128,0.009737778041097853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,5120,32,0.0025395556456512874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,64,0.009411555197503831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,65536,0.08708533313539292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,5120,32,0.00870666652917862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,51200,0.06889600223965116
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,65536,0.06646222538418241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,16384,0.028166221247778997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,51200,0.07925511068767972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,12288,0.02156266735659705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,16384,0.031239112218221027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,12288,0.028860443168216284
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,10240,0.018280888597170513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,8192,0.015600888265503777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,10240,0.025157334076033697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,7168,0.014555555250909595
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,8192,0.02422222163942125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,6144,0.01310222182008955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,7168,0.022292445103327434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,5120,0.012072888513406118
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,6144,0.01980355547534095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,4096,0.010943111446168689
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,5120,0.01735466718673706
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,3584,0.010585777461528778
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,4096,0.01683644453684489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,3584,0.014854222536087036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,3072,0.010946667028797997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,2560,0.009712888962692684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,3072,0.014163555370436775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,2048,0.007334222396214803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,2560,0.013463111387358772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,1536,0.006664888726340399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,2048,0.012780444489585029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,1024,0.005342222336265776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,1536,0.01274311128589842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,768,0.004911111046870549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,1024,0.011168888873524137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,512,0.003936888857020272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,768,0.012116444607575735
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,512,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,256,0.003185777821474605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,128,0.0029564443975687027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,256,0.010783111055692038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,64,0.0025457777082920074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,128,0.009727111293209923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,4096,32,0.0028248888750871024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,64,0.008978666530715095
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,65536,0.07616711325115628
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,4096,32,0.009031111167536842
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,51200,0.06156444549560547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,65536,0.06273066997528076
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,16384,0.023732443650563557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,51200,0.06881688700781928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,12288,0.01886666648917728
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,16384,0.030629333522584703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,10240,0.016520889268981088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,10240,0.024067555864651997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,12288,0.028904000918070476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,8192,0.014511111709806653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,7168,0.013448889056841532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,8192,0.02369333306948344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,7168,0.02100444502300686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,6144,0.012183999849690331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,5120,0.011556444068749746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,6144,0.018589332699775696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,4096,0.010224888722101847
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,5120,0.01733955575360192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,4096,0.015511999527613321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,3584,0.010238221950001186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,3072,0.009343999955389235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,3584,0.014562666416168213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,2560,0.008661333057615492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,3072,0.014369777507252164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,2048,0.007050666544172499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,2560,0.013122667041089801
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,1536,0.00629066675901413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,2048,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,1024,0.0052800000541739995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,1536,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,768,0.004285333471165763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,768,0.011447110937701331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,512,0.0053137776752312975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,1024,0.012151999606026543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,256,0.003148444410827425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,128,0.0028284444577164124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,512,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,128,0.009638222555319468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,64,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3584,32,0.002550222186578645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,64,0.00869599978129069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,65536,0.06748977634641859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,32,0.009013333254390294
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,65536,0.059805333614349365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,51200,0.05445955528153313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,51200,0.07038578059938219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,16384,0.0210515558719635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3584,256,0.011409777734014722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,12288,0.016972444123691983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,16384,0.02927466564708286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,10240,0.015235554840829639
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,12288,0.025651555922296312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,8192,0.013430222041077085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,10240,0.022831110490692988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,7168,0.012862222062216865
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,8192,0.021381333470344543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,6144,0.011402666568756104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,6144,0.018260444204012554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,5120,0.010407999985747868
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,5120,0.01575911045074463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,4096,0.009648000200589498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,7168,0.020181333025296528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,3584,0.009097778134875828
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,3584,0.01421955558988783
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,3072,0.008053333395057255
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,3072,0.013543999857372709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,2560,0.0070471109615431885
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,2560,0.013111111190583972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,2048,0.0063324446479479475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,4096,0.014512888259357877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,2048,0.012471111284361945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,1536,0.006330666856633292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,1024,0.005629333357016246
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,1536,0.011136000355084738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,768,0.004654222064548069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,1024,0.011776000261306763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,768,0.01112711098459032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,512,0.004271111140648524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,256,0.0035173334181308746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,512,0.010436444646782346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,128,0.002850666642189026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,256,0.01145511120557785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,64,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,128,0.009658666948477427
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,3072,32,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,64,0.009220444493823582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,3072,32,0.00867377801073922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,65536,0.057998220125834145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,51200,0.04726666543218824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,65536,0.057481778992546924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,16384,0.020264888803164165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,51200,0.06336444616317749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,16384,0.027863999207814533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,12288,0.01499644418557485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,10240,0.013924444715181986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,12288,0.024530667397711012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,8192,0.012109333442317115
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,10240,0.02207999924818675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,7168,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,8192,0.02070488863521152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,6144,0.010451555252075195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,7168,0.01833599971400367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,5120,0.010096888575288985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,6144,0.016705777910020616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,4096,0.009171555439631144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,5120,0.014343111051453484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,4096,0.013800889253616333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,3584,0.008647110727098253
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,3072,0.008040889269775814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,3584,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,2560,0.0069288888739215
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,3072,0.01313511116637124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,2048,0.005944889038801193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,2560,0.012483555409643384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,1536,0.00555644432703654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,2048,0.012434666355450949
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,1024,0.004920888692140579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,1536,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,768,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,1024,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,512,0.0038888889054457345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,768,0.011101333631409539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,256,0.0032462223122517266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,512,0.011083555718262991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,256,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,64,0.0028719999310043124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,128,0.009381333159075843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2560,32,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,64,0.009352000223265754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,65536,0.04732000165515476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2560,32,0.008370666868156856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,51200,0.03875999980502658
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,65536,0.05587289068433973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,16384,0.015393777026070489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,51200,0.060198220941755504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,12288,0.01293333371480306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,16384,0.024134222004148696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,10240,0.011973333027627734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,12288,0.02271644439962175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,8192,0.01110222190618515
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,10240,0.020248000820477802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,8192,0.018739556272824604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,7168,0.010297777752081553
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,7168,0.01684266659948561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,6144,0.009667555491129557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,6144,0.014843554960356818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,5120,0.008966222405433655
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,5120,0.013854222165213691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,4096,0.007703999678293864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,4096,0.013582222163677216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,3584,0.007349333001507654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,3584,0.013191111385822296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,3072,0.0069617778062820435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,3072,0.01313244468635983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,2560,0.006286222073766921
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,2560,0.01276533305644989
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,2048,0.008081778056091731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,2048,0.012368888490729861
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,1536,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,1536,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,1024,0.005249777601824866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,768,0.004575110971927643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,1024,0.01183733344078064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,512,0.003968888686762916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,768,0.011705777711338468
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,256,0.0032595555401510666
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,512,0.011155555645624796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,128,0.00313688897424274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,256,0.011051555474599203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,128,0.009731555150614845
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,2048,32,0.002941333378354708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,64,0.009378666679064432
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,65536,0.03637689020898607
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,2048,32,0.008985777695973715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,51200,0.02981333269013299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,65536,0.05423644516203138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,16384,0.013794666363133324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,51200,0.049039999643961586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,12288,0.011394666300879585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,16384,0.02368444369898902
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,10240,0.01311644415060679
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,12288,0.020970667401949566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,10240,0.018539556198649935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,8192,0.011737777955002256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,7168,0.011108444796668159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,8192,0.015878222054905362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,6144,0.008859555754396651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,7168,0.015549333559142219
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,5120,0.007260444263617198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,6144,0.014544000228246054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,4096,0.007423111134105259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,5120,0.014107555150985718
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,3584,0.007085333267847697
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,4096,0.013470222552617392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,3072,0.006292444550328785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,3584,0.01344622257683012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,2560,0.005711110929648082
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,3072,0.01312888910373052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,2048,0.005697777701748743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,2560,0.012897777888509961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,2048,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,1536,0.005255110975768831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,1024,0.00424177779091729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,1536,0.011741333537631564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,768,0.003896000070704354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,1024,0.011672000090281168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,512,0.003938666648334927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,768,0.011392888923486074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,256,0.0032257777121331957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,512,0.01109155515829722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,128,0.003134222287270758
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,256,0.011075555450386472
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,64,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,128,0.010093332992659675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,64,0.009355555805895064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1536,32,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,65536,0.02622577713595496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1536,32,0.00903733323017756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,51200,0.02102222210831112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,65536,0.05143288771311442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,16384,0.014117333624098035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,51200,0.04286844531695048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,12288,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,16384,0.021319111188252766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,10240,0.01072088877360026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,12288,0.017903111047214933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,8192,0.009696000152164036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,10240,0.01718488832314809
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,7168,0.008727110922336578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,8192,0.015569777952300178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,6144,0.0069724445541699724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,7168,0.01555377741654714
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,5120,0.006665777828958299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,6144,0.014488889111412896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,4096,0.007281777759393056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,5120,0.01421511173248291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,4096,0.013250666360060373
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,3584,0.0069368887278768755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,3072,0.0063662222690052455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,3584,0.013089777694808112
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,2560,0.005592888842026393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,3072,0.013136000268989138
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,2560,0.012818666795889536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,2048,0.006315555423498154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,2048,0.012494222157531314
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,1536,0.0053324442770746015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,1536,0.012077333198653327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,1024,0.004300444490379757
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,1024,0.011766222616036734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,768,0.003935111065705617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,768,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,512,0.003690666622585721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,512,0.011110222174061669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,256,0.0034328889515664843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,256,0.011398221883508893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,128,0.0032177778581778207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,128,0.010077333284748925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,64,0.0029271110478374693
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,64,0.009073778159088558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,1024,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,65536,0.021372444099850122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,1024,32,0.009043555292818282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,51200,0.0177831103404363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,65536,0.05067111055056254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,16384,0.012097777591811286
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,51200,0.04258044560750326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,12288,0.010074666804737514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,16384,0.020968000094095867
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,10240,0.009337777892748514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,12288,0.01794399983353085
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,8192,0.00868622213602066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,10240,0.01661066710948944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,7168,0.00832088871134652
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,8192,0.01568177839120229
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,6144,0.007009777757856581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,7168,0.01460000044769711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,6144,0.014510222607188754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,5120,0.00665866666369968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,4096,0.006955555743641323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,5120,0.013759111364682516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,3584,0.00664444433318244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,4096,0.012847111456924014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,3072,0.006309333360857434
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,3584,0.013129777378506131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,2560,0.006271111054552927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,3072,0.012424888710180918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,2048,0.004975999808973736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,2560,0.012497777740160624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,2048,0.011487111449241638
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,1536,0.004967999955018361
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,1024,0.004557333472702238
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,1536,0.011090666883521609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,768,0.0042239998777707415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,1024,0.011607999602953592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,512,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,768,0.010811555716726514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,256,0.0031991110493739447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,512,0.011119999819331698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,128,0.002895999906791581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,256,0.01070666644308302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,64,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,128,0.010382222632567087
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,768,32,0.002891555635465516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,64,0.009334222310119206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,65536,0.023007111416922674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,768,32,0.008984888593355814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,51200,0.018906666172875296
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,65536,0.050272888607449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,16384,0.010797333386209277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,51200,0.04244533181190491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,12288,0.008385777473449707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,16384,0.019688000281651814
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,10240,0.007693332930405934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,12288,0.01796800063716041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,8192,0.007811555431948767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,10240,0.016286222471131217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,7168,0.007352888584136963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,8192,0.015828443898095023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,6144,0.007294221884674496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,7168,0.014471999473041959
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,5120,0.006640888750553131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,6144,0.014559111661381192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,4096,0.006994666324721442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,5120,0.013178666432698568
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,3584,0.006592888798978593
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,4096,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,3072,0.006615110983451207
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,3584,0.012157333393891653
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,2560,0.006032889087994893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,3072,0.012885332935386233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,2048,0.005653333332803514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,2560,0.012112889024946423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,2048,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,1536,0.004600888739029567
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,1024,0.00388711111413108
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,1536,0.011764444410800934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,768,0.0035706667436493766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,1024,0.010715555813577441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,512,0.003220444545149803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,768,0.011711999773979187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,256,0.0032044444233179092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,512,0.010524444282054901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,256,0.010996444357766045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,128,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,128,0.010120000276300643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,64,0.0028826666788922418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,64,0.009351111120647853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,512,32,0.0028817777832349143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,512,32,0.00903733323017756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,65536,0.015531554818153381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,65536,0.04629688792758518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,51200,0.012786666552225748
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,16384,0.00851466672288047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,51200,0.03906666570239597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,12288,0.008006222546100616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,16384,0.01961688862906562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,10240,0.007315555380450354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,12288,0.01760088900725047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,10240,0.015846222639083862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,8192,0.006989333364698622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,7168,0.006252444452709622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,8192,0.015108444624476962
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,6144,0.006262222097979651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,7168,0.014519999424616495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,5120,0.00600888869828648
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,6144,0.013535999589496188
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,4096,0.005974222388532426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,5120,0.013140444126394061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,3584,0.005975111077229182
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,4096,0.012747555143303342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,3072,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,3584,0.01237066669596566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,2560,0.005575110928879843
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,3072,0.012103111379676394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,2048,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,2560,0.01203644441233741
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,1536,0.004265777766704559
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,2048,0.011427555647161273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,1024,0.003589333345492681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,1536,0.011232888533009423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,1024,0.01053066634469562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,768,0.003519111209445529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,512,0.003370666669474708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,768,0.010805333654085795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,256,0.002903999967707528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,512,0.010420444111029306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,128,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,256,0.01037688884470198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,64,0.0028088889602157804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,128,0.010047111246320937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,256,32,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,64,0.00867911097076204
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,65536,0.013176889055305056
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,256,32,0.009034666750166152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,51200,0.01182577759027481
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,51200,0.03818044397566054
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,16384,0.007595555649863348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,16384,0.019556444552209642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,12288,0.007220444579919179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,12288,0.01738933391041226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,10240,0.006970666348934174
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,10240,0.016280000408490498
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,65536,0.04538133409288195
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,8192,0.0063164445261160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,7168,0.006291555447710886
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,8192,0.015113777584499784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,6144,0.005724444571468566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,7168,0.014528888795110913
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,5120,0.005869333528810077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,6144,0.013432888521088494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,4096,0.0052622221410274506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,5120,0.013068444199032254
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,3584,0.005560889012283749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,4096,0.012445333103338877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,3072,0.005608888963858287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,3584,0.012206222448084088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,2560,0.005581333405441708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,3072,0.012111110819710625
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,2048,0.005606222069925732
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,2560,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,1536,0.004214222232500712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,2048,0.011424000064531961
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,1024,0.0038444445365005066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,1536,0.01108088923825158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,768,0.0034622223012977173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,1024,0.010379555324713389
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,512,0.003181333343187968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,768,0.010925333533022137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,256,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,512,0.010249777800507015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,128,0.0028213332924577924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,256,0.011046222514576383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,64,0.002566222308410539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,128,0.009288888838556077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,128,32,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,64,0.009325332939624786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,65536,0.010410666465759277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,128,32,0.00829777783817715
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,51200,0.009024889104896123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,65536,0.04556177722083198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,51200,0.03775022096104092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,16384,0.00665155549844106
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,16384,0.01940444442960951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,12288,0.0059075554211934405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,12288,0.017369776964187622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,10240,0.005791999813583162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,10240,0.016307555966907077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,8192,0.005893333504597346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,8192,0.014852444330851236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,7168,0.005560889012283749
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,7168,0.014482667048772177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,6144,0.005636444522274866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,5120,0.006024000131421619
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,6144,0.014056889547242058
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,4096,0.0052382221652401825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,5120,0.013286222186353473
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,4096,0.012476444244384766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,3584,0.006031999985376994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,3072,0.005666666560702854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,3584,0.01219466659757826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,2560,0.005992888990375731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,3072,0.01255022237698237
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,2048,0.006136888845099344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,2560,0.012089778151777057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,1536,0.0042071110672420925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,2048,0.012119111087587146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,1024,0.003871110992299186
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,1536,0.011056000159846412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,768,0.0031866667171319327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,1024,0.011445333560307821
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,512,0.0032497776879204642
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,768,0.010866666833559671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,256,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,512,0.01109333336353302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,128,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,256,0.01069777790043089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,64,0.002485333424475458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,128,0.00980711148844825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,64,32,0.002986666642957263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,64,0.009324444664849175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,65536,0.008993777963850234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,64,32,0.009010666774378883
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,51200,0.008336000144481659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,65536,0.04554222358597649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,16384,0.005917333480384614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,51200,0.03839288817511664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,12288,0.005950222412745158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,16384,0.01960533360640208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,10240,0.005608888963858287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,12288,0.01753422286775377
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,8192,0.005233777893914117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,10240,0.016179554992251925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,8192,0.015527110960748462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,7168,0.005587555468082428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,6144,0.005573333137565189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,7168,0.014779556128713818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,5120,0.005239999956554837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,6144,0.014101333088344999
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,4096,0.005296000176005893
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,5120,0.01314222233162986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,3584,0.005625777774386936
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,4096,0.012740444805887012
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,3072,0.0052800000541739995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,3584,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,2560,0.005597333527273602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,3072,0.012220444778601328
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,2560,0.012087999946541257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,2048,0.011738667057620155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,1536,0.004602666530344221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,1536,0.01145333300034205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,1024,0.003918222255176968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,1024,0.011390222443474663
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,768,0.003544888976547453
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,768,0.011412444214026133
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,2048,0.0052844443255000645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,512,0.0032035555276605818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,256,0.0029235554652081597
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,512,0.01037599974208408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,256,0.011065777805116443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,128,0.0025271111064487034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,64,0.002567111204067866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,128,0.010044444766309526
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,2,32,32,0.002514666567246119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,64,0.009373333719041612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,2,32,32,0.009024889104896123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,51200,0.5599093437194824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,51200,0.9591066572401258
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,16384,0.18058399359385172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,16384,0.3319351143307156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,12288,0.2408346600002713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,10240,0.20301600297292074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,12288,0.15403555499182806
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,8192,0.16501066419813368
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,8192,0.11965866883595784
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,7168,0.14554311169518366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,7168,0.10931822326448228
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,6144,0.12775111198425293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,5120,0.10806488990783691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,6144,0.09382933378219604
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,4096,0.0888142254617479
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,5120,0.08179733488294813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,10240,0.13440799713134766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,4096,0.06761599911583795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,3584,0.07226488987604777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,3072,0.06232711341645983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,3584,0.06518844763437907
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,2560,0.052785777383380465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,2560,0.05010044574737549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,3072,0.056695110268063016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,2048,0.042824000120162964
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,1536,0.03311377763748169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,2048,0.04124088750945197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,1536,0.03489333391189575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,1024,0.02351733379893833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,768,0.018602665927675035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,1024,0.02630222174856398
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,512,0.01365600029627482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,768,0.022720000810093347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,256,0.00888888869020674
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,256,0.01513066722287072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,128,0.006007110906971826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,128,0.012472888661755456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,64,0.005649777750174205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,512,0.01901511185699039
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,65536,32,0.005678222411208683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,64,0.012170666621790992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,65536,32,0.012200888660218982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,65536,0.5349289046393501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,65536,1.0754533343844943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,51200,0.8427599800957574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,51200,0.44846312204996747
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,16384,0.2685573365953234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,16384,0.14897599485185412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,12288,0.19702577590942383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,10240,0.16650399896833631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,12288,0.1269013351864285
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,8192,0.13657511605156794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,10240,0.1086382203631931
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,8192,0.09733333190282185
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,7168,0.12000000476837158
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,6144,0.1057902243402269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,7168,0.08914399809307522
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,5120,0.08993244171142578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,6144,0.07707555426491632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,4096,0.0731840001212226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,5120,0.06899022393756442
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,4096,0.057393776045905225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,3584,0.0655884411599901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,3072,0.057145780987209745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,3584,0.05560977591408623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,2560,0.04415555463896858
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,3072,0.04955288767814636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,2048,0.037704888317320086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,2560,0.042633778519100614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,1536,0.028392887777752344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,2048,0.035657776726616755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,1536,0.028838220569822524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,1024,0.02286311156219906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,768,0.015779554843902588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,768,0.020364445116784837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,512,0.011368888947698804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,512,0.01756888959142897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,256,0.007387555307812161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,1024,0.020756444997257657
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,256,0.013459555804729462
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,128,0.003915555361244413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,64,0.003588444242874781
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,128,0.011752888560295105
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,51200,32,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,64,0.01148088855875863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,51200,32,0.01144977741771274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,65536,0.3277555571662055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,51200,0.2552693419986301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,16384,0.09395111269421047
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,51200,0.17115467124515107
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,12288,0.06744355625576444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,16384,0.08139289087719388
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,10240,0.0577404432826572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,12288,0.05884355306625366
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,8192,0.048390222920311816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,10240,0.052871998813417226
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,7168,0.03814222084151374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,65536,0.22982401318020293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,8192,0.044256889157825045
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,6144,0.03329955538113912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,7168,0.042211555772357516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,5120,0.028435554769304063
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,4096,0.023608888188997906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,6144,0.03648089038001166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,5120,0.03250044584274292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,3584,0.020564445190959506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,3072,0.017865777015686035
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,4096,0.026762665973769292
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,3584,0.026139555705918208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,2560,0.016024889217482675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,2048,0.013160000244776407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,3072,0.023769777682092454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,2560,0.021900445222854614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,2048,0.019945777124828763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,1024,0.008346666892369589
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,1536,0.017337777548366122
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,1024,0.015383111106024848
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,768,0.007037333316273159
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,768,0.013895110951529609
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,512,0.004922666483455234
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,512,0.013091555900043912
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,256,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,1536,0.01070755554570092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,128,0.0029128889242808023
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,256,0.011104889214038849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,64,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,128,0.009764444496896531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,64,0.009744000103738572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,16384,32,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,16384,32,0.009744000103738572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,65536,0.23923643430074057
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,51200,0.18789954980214438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,65536,0.1926168865627713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,16384,0.0714284446504381
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,51200,0.1463466617796156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,16384,0.060434665944841176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,12288,0.051848000950283475
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,10240,0.044305778212017484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,12288,0.050660444630516895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,8192,0.03751911057366265
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,10240,0.044044444958368935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,8192,0.03852622376547919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,7168,0.03282311227586534
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,6144,0.029976887835396662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,7168,0.03478844298256768
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,5120,0.025809778107537165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,6144,0.030743112166722614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,4096,0.02207022243075901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,5120,0.02759644389152527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,3584,0.019567110472255282
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,4096,0.024697777297761705
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,3072,0.017715555098321702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,3584,0.024175999893082514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,2560,0.015811555915408664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,3072,0.021831999222437542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,2048,0.010586666564146677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,2560,0.020627554919984605
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,1536,0.00871022211180793
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,2048,0.01776355504989624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,1536,0.016958221793174744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,1024,0.0072266666425599
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,768,0.005716444303592046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,1024,0.014191110928853353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,512,0.00360355567600992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,768,0.013341333303186627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,512,0.012467555701732635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,256,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,128,0.0028684443483750024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,256,0.011066666907734342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,64,0.0028986665937635633
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,128,0.00980355590581894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,64,0.009764444496896531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,12288,32,0.0028488888508743713
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,12288,32,0.009691555466916826
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,65536,0.20031732983059355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,51200,0.1581475602255927
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,65536,0.16976000203026664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,16384,0.061271111170450844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,51200,0.1294835540983412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,12288,0.0429057776927948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,16384,0.05301688777075874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,10240,0.03746844331423441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,12288,0.044128888183169894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,8192,0.0303439994653066
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,10240,0.03872799873352051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,7168,0.02750844425625271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,7168,0.029808001385794744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,8192,0.03251288996802436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,6144,0.024675556355052527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,5120,0.02219999995496538
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,6144,0.026782222919993933
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,4096,0.01850222216712104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,4096,0.02134311033619775
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,3584,0.017497777938842773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,3584,0.021015110943052504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,3072,0.01610488858487871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,5120,0.02436888880199856
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,3072,0.019323555959595572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,2560,0.012058667010731168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,2048,0.010512000156773461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,2560,0.01793688866827223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,2048,0.0170133329100079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,1536,0.00942488925324546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,1024,0.007321777443091075
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,1536,0.015573332707087198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,768,0.00563822231358952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,1024,0.012544000314341651
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,512,0.004552888787455029
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,768,0.012746666868527731
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,256,0.002975111206372579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,256,0.010743111371994019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,128,0.0028533333291610083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,128,0.009744889206356471
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,64,0.002532444480392668
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,64,0.009727111293209923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,10240,32,0.002867555452717675
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,512,0.0122417774465349
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,10240,32,0.009697777529557547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,65536,0.16164888275994196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,65536,0.12090488274892171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,51200,0.12844710879855686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,51200,0.09344800313313802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,16384,0.04989333285225762
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,12288,0.03282489048110114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,16384,0.04055111275778876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,10240,0.02887644370396932
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,12288,0.0343413319852617
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,8192,0.0233297778500451
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,10240,0.030253334177864924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,7168,0.021353777911927965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,8192,0.02684266699684991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,7168,0.024698666400379602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,6144,0.018617777360810172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,5120,0.016221332881185744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,6144,0.023016888234350417
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,4096,0.013804444836245643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,5120,0.020970667401949566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,3584,0.01237511138121287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,4096,0.018232888645595975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,3072,0.011350221931934357
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,3584,0.018945778409639995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,2560,0.009840889109505547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,3072,0.017198221551047433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,2048,0.008605333666006723
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,2560,0.017177777157889474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,1536,0.007105777661005656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,2048,0.01458666721979777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,1024,0.005376889059940974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,1536,0.013804444836245643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,768,0.0042568888101312844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,1024,0.012080888781282636
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,512,0.0038951109680864546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,768,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,256,0.0031999999450312722
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,512,0.011773332953453064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,128,0.0028524444335036804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,256,0.011032000184059143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,128,0.009375111096435124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,64,0.0028373334142896864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,8192,32,0.0028311111446883944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,64,0.009374221993817223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,8192,32,0.010046222143703038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,65536,0.14230933454301622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,51200,0.1129973332087199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,65536,0.12206044461992051
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,16384,0.04215911030769348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,51200,0.08485777510537042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,12288,0.03218844532966614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,16384,0.044714666075176664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,12288,0.033815999825795494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,10240,0.02840533190303379
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,8192,0.023790222075250413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,10240,0.02903199858135647
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,7168,0.021128889587190416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,8192,0.026512000295850966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,6144,0.017302221722073026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,7168,0.024189333120981853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,5120,0.015106666419241162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,6144,0.022664889693260193
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,4096,0.012749333348539142
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,5120,0.019931554794311523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,3584,0.011413333316644033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,4096,0.018862222631772358
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,3584,0.018220444520314533
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,3072,0.010488000180986194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,2560,0.009139555195967356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,3072,0.016921778519948322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,2560,0.016908443636364408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,2048,0.008220444122950235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,1536,0.006960888703664144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,2048,0.014481777946154276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,1024,0.004519999855094486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,1536,0.013086222112178802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,768,0.0038657776183552216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,1024,0.012367111113336353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,512,0.0038746665749284956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,768,0.012135110795497894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,256,0.003006222140457895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,512,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,128,0.002924444360865487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,256,0.010419555836253695
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,64,0.002499555548032125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,128,0.01015644437736935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,7168,32,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,64,0.009402666654851701
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,65536,0.1225351095199585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,7168,32,0.010058666268984476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,51200,0.09915644592709011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,65536,0.10252888997395833
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,16384,0.039247112141715154
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,51200,0.07837777667575412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,12288,0.029530667596393164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,16384,0.03675466775894165
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,10240,0.025755556093321905
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,12288,0.030406223403082952
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,8192,0.021031111478805542
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,10240,0.027133334014150832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,7168,0.01960533360640208
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,8192,0.025026667449209426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,6144,0.017431111799346078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,7168,0.023123555713229712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,5120,0.0158933334880405
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,6144,0.021557332740889654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,4096,0.011585777832402123
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,5120,0.019136889113320243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,3584,0.01072177787621816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,4096,0.018216888109842937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,3072,0.009431111315886179
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,3584,0.017644445101420086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,2560,0.00869777798652649
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,3072,0.016915554801623028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,2048,0.007502222226725684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,2560,0.014863111906581454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,1536,0.006279110908508301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,2048,0.014488000008794995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,1024,0.004254222330119875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,1536,0.01313511116637124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,768,0.0038817777401871155
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,1024,0.012416000167528788
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,512,0.003571555432346132
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,768,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,256,0.0031742221779293488
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,512,0.012072000238630505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,128,0.00305244450767835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,256,0.010423111418883005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,64,0.0028391112056043413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,128,0.010084444450007545
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,6144,32,0.0029297777348094513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,64,0.009420444567998251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,65536,0.10315555334091187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,6144,32,0.010029333333174387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,51200,0.08377689123153687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,65536,0.09054666757583618
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,16384,0.03392622205946181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,51200,0.06951110892825656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,12288,0.02527911133236355
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,16384,0.03330577744377984
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,10240,0.021759111020300124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,12288,0.028536889288160536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,8192,0.019221333993805777
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,10240,0.025109334124459162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,7168,0.016846223009957206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,8192,0.023779556155204773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,6144,0.015840000576443143
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,7168,0.021567111214001972
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,5120,0.01370844410525428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,6144,0.020319110817379422
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,4096,0.010659555594126383
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,5120,0.01812355551454756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,3584,0.009758222434255812
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,4096,0.017753778232468497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,3072,0.00868266655339135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,3584,0.01664977769056956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,2560,0.007920000288221572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,3072,0.01590222285853492
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,2048,0.007012444237867991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,2560,0.01382844481203291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,1536,0.00507733349998792
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,2048,0.013769778112570444
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,1024,0.004224888980388641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,1536,0.012498666842778524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,768,0.003946666502290302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,1024,0.012038222617573209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,512,0.003564444267087512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,768,0.011775111158688864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,256,0.0031946665710873077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,512,0.011756444142924415
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,128,0.0029004443850782183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,256,0.01037688884470198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,64,0.002854222224818336
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,128,0.01017599966790941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,5120,32,0.002814222127199173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,64,0.009366222553782994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,65536,0.08455910947587754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,5120,32,0.010048000348938836
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,51200,0.06916977961858113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,65536,0.08582311206393772
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,16384,0.02793244520823161
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,51200,0.06303377946217854
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,12288,0.02163999941613939
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,16384,0.03154044349988302
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,10240,0.01831733352608151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,12288,0.02681333323319753
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,8192,0.01553066737122006
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,10240,0.02425066630045573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,7168,0.014519110321998596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,8192,0.022403554783927068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,6144,0.01311822235584259
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,7168,0.02113688819938236
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,5120,0.011955555942323474
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,6144,0.019633778267436557
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,4096,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,5120,0.017873777283562552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,3584,0.011055111057228513
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,4096,0.016934222645229764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,3072,0.01038933296998342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,3584,0.015967110792795818
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,3072,0.014476444986131458
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,2560,0.009669333696365356
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,2560,0.013843555417325763
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,2048,0.008359111017651027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,2048,0.012841777669058906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,1536,0.006955555743641323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,1536,0.012486221889654795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,1024,0.005487999982304043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,1024,0.01131199962562985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,768,0.004568888909286923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,768,0.011771555576059552
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,512,0.003940444439649582
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,512,0.011396444506115384
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,256,0.003261333331465721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,128,0.003040888864133093
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,256,0.010431999961535135
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,64,0.002696000039577484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,128,0.009409777820110321
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,4096,32,0.0029057777590221832
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,64,0.009340444372759925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,65536,0.07561066415574816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,4096,32,0.009171555439631144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,51200,0.06122399701012505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,65536,0.08052888843748304
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,16384,0.02437866727511088
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,51200,0.06118933359781901
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,12288,0.018977777825461496
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,16384,0.03052799900372823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,10240,0.01678400072786543
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,12288,0.02552177839808994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,8192,0.014348444011476306
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,10240,0.023358222511079576
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,7168,0.013335111240545908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,8192,0.020954666866196528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,6144,0.012879110872745514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,7168,0.0203493336836497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,5120,0.011057777537239922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,6144,0.018400889303949144
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,4096,0.010469333165221743
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,5120,0.017443555924627516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,3584,0.010464888479974536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,4096,0.015288889408111572
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,3072,0.012477333347002665
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,3584,0.015365334020720588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,2560,0.008662222160233391
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,3072,0.013879111243618859
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,2048,0.007651555869314406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,2560,0.013458666702111563
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,1536,0.006264888991912206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,2048,0.012907555533779992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,1024,0.0052844443255000645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,1536,0.01235466698805491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,768,0.0045422220395671
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,1024,0.011755555868148804
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,512,0.005607999861240387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,768,0.011744889120260874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,256,0.00291377781993813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,256,0.01015733347998725
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,128,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,128,0.009411555197503831
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,64,0.002534222271707323
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,512,0.011399110986126794
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,64,0.009445333646403419
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3584,32,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,65536,0.06588088803821139
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3584,32,0.009257777697510189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,51200,0.05484355489412943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,65536,0.07442133294211493
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,16384,0.020827554994159274
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,51200,0.0588408907254537
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,12288,0.016937777400016785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,16384,0.027436445156733196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,10240,0.015245333313941956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,12288,0.024499555428822834
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,8192,0.013253333667914072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,10240,0.021149333980348375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,7168,0.012210666305489011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,8192,0.019996444384257
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,7168,0.018562666243977018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,6144,0.011094222466150919
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,5120,0.010378667049937779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,6144,0.017512000269360013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,4096,0.009568889107969072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,5120,0.015999111864301894
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,3584,0.00908799966176351
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,4096,0.014141332772043018
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,3072,0.008343111309740279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,3584,0.014496000276671516
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,2560,0.007872000336647034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,3072,0.013442666994200813
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,2048,0.00638488887084855
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,2048,0.01261333293384976
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,1536,0.006631999793979857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,2560,0.013013333082199097
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,1024,0.005240889059172736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,1536,0.012513777448071374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,768,0.004920888692140579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,1024,0.011044444309340583
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,512,0.003948444293604957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,768,0.011768889096048145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,512,0.0114702218108707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,256,0.0033155555526415506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,256,0.010050666828950247
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,128,0.002847111059559716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,128,0.009350222018029954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,64,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,64,0.009004444711738164
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,3072,32,0.0028702221396896574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,3072,32,0.00904888908068339
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,65536,0.05760177638795641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,65536,0.07166488965352376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,51200,0.04702755477693346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,51200,0.05627644724316067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,16384,0.020245333512624104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,16384,0.026467555099063452
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,12288,0.015415110521846347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,10240,0.014127999544143677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,12288,0.022655111220147874
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,8192,0.011774222056070963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,10240,0.021951999929216173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,7168,0.011397333608733283
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,8192,0.019072888625992667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,6144,0.010579555398888059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,7168,0.01901955571439531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,5120,0.009741333623727163
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,6144,0.01614755557643043
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,4096,0.009321778184837764
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,5120,0.015150222513410779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,3584,0.008336889247099558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,4096,0.013766222529941134
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,3072,0.008358221915033128
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,3584,0.013758222262064615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,2560,0.007631111476156447
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,3072,0.012801777985360889
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,2048,0.006620444357395172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,2560,0.012472000386979846
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,1536,0.0063733334342638654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,2048,0.012130666938092975
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,1024,0.005236444373925527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,1536,0.011727111207114326
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,768,0.004551110996140374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,1024,0.011481777661376528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,512,0.003859555555714501
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,768,0.010732444624106089
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,256,0.0033928888539473214
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,512,0.011081777513027191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,128,0.002871111035346985
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,256,0.00978577799267239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,128,0.009285333255926767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,64,0.002858666703104973
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2560,32,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,64,0.009022222624884712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2560,32,0.00867555538813273
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,65536,0.04717422193951077
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,51200,0.03922755519549052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,65536,0.07175288597742717
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,16384,0.015404444601800708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,51200,0.05008177624808418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,12288,0.01274400038851632
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,16384,0.02727288835578495
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,10240,0.011682666838169098
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,12288,0.02090666691462199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,8192,0.01069155583779017
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,10240,0.020107555720541213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,7168,0.010314666562610203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,8192,0.01756355497572157
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,6144,0.009404444032245213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,7168,0.016527111331621807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,6144,0.014848889576064216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,5120,0.008819555242856344
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,4096,0.007996444072988298
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,5120,0.014104889498816596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,3584,0.007309333317809635
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,4096,0.01312355531586541
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,3072,0.0069439998931354946
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,3584,0.013067555924256643
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,2560,0.006582222051090664
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,3072,0.011854222251309289
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,2048,0.007634667058785756
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,2560,0.012792888614866467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,1536,0.006578666468461354
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,2048,0.011385777758227454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,1024,0.005252444495757421
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,1536,0.012099555797047086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,768,0.00453333349691497
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,1024,0.010673777924643623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,512,0.0038897775941424896
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,768,0.011353777514563667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,256,0.0032017777363459268
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,512,0.010672888822025724
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,128,0.0031884445084465873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,256,0.010070222119490305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,64,0.0028808888875775864
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,64,0.008991111483838823
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,2048,32,0.0028293333533737394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,32,0.008992888861232335
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,65536,0.035184890031814575
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,65536,0.05807822280459934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,51200,0.028424888849258423
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,2048,128,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,51200,0.051820443736182324
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,16384,0.013529777526855469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,16384,0.02407466702991062
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,12288,0.011155555645624796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,12288,0.02088000045882331
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,10240,0.013966222604115805
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,10240,0.017722666263580322
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,8192,0.0123848890264829
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,8192,0.015601777368121676
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,7168,0.011428444749779172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,7168,0.015127999915017022
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,6144,0.008580444587601556
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,6144,0.014134221606784396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,5120,0.007348444726732042
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,5120,0.013808888693650564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,4096,0.007146666447321574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,4096,0.013108444710572561
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,3584,0.006806222101052602
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,3584,0.01275022245115704
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,3072,0.006813333266311222
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,2560,0.005903999838564131
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,3072,0.011769777370823754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,2048,0.006284444282452266
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,2560,0.012774222426944308
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,1536,0.0052639999323421055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,2048,0.011095999843544431
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,1024,0.004570666700601578
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,1536,0.012140444583363004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,1024,0.010695111420419482
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,768,0.003912888881233004
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,512,0.003535110917356279
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,768,0.010771555205186209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,256,0.003228444399105178
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,512,0.012101333174440594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,128,0.0028453332682450614
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,256,0.009670221971140968
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,64,0.0028417776856157514
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,128,0.010073777702119615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,64,0.008398222426573435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1536,32,0.008384889198674096
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,65536,0.02642400066057841
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,65536,0.05415999889373779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,51200,0.02092355489730835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,51200,0.04486755530039469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,16384,0.013688888814714221
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1536,32,0.003015111097031169
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,16384,0.02221244408024682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,12288,0.012481778032249875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,10240,0.011485333244005838
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,12288,0.01754666699303521
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,8192,0.010105777945783403
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,10240,0.016972444123691983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,7168,0.0090844440791342
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,8192,0.01552977826860216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,6144,0.0075146663520071245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,7168,0.01550222271018558
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,6144,0.014194667339324951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,5120,0.006953777538405524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,4096,0.007061333292060428
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,5120,0.013743999931547375
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,3584,0.006424888968467712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,4096,0.013753777576817406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,3584,0.012840000291665396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,3072,0.006272888845867581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,2560,0.005626666463083691
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,3072,0.012079111403889127
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,2048,0.007313778003056844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,2560,0.012818666795889536
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,1536,0.006430222342411677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,2048,0.011134222149848938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,1024,0.004958222309748332
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,1536,0.012044444680213928
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,1024,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,768,0.004574222283230888
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,512,0.003867555409669876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,768,0.010770666930410596
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,256,0.0032791110376516977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,256,0.009468444519572789
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,128,0.002941333378354708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,512,0.011891555454995898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,64,0.002838222309947014
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,128,0.008683555656009251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,1024,32,0.0028168888141711554
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,64,0.010098666780524785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,1024,32,0.00831911133395301
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,65536,0.021003555920388963
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,51200,0.017695999807781644
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,65536,0.053871999184290566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,16384,0.012118221984969245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,51200,0.04255733225080702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,12288,0.010328888893127441
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,16384,0.021018667353524104
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,10240,0.009390222529570261
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,12288,0.01828266680240631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,8192,0.00869955536392
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,10240,0.016136889656384785
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,7168,0.008374222450786168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,8192,0.015821332732836407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,7168,0.014540443817774454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,6144,0.007116444408893585
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,6144,0.014840000205569796
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,5120,0.00657066661450598
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,5120,0.014180445008807711
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,4096,0.006958222223652734
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,4096,0.01383733335468504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,3584,0.006647111227114995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,3584,0.0138568886452251
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,3072,0.006331555545330048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,2560,0.006487999939256244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,3072,0.012885332935386233
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,2048,0.005122666557629903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,2560,0.011791110866599612
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,2048,0.012305777933862476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,1536,0.005444444302055571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,1024,0.004319111092223061
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,1536,0.01201866649919086
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,768,0.003945777813593547
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,1024,0.010718222293588849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,512,0.0036142224238978494
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,768,0.011777777638700275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,256,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,512,0.011077333655622272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,128,0.0028693332440323303
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,256,0.010752889017264048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,64,0.0028853333658642242
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,128,0.009375111096435124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,768,32,0.0028595555987623003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,64,0.00999644481473499
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,65536,0.022713777091768052
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,768,32,0.009062222308582729
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,51200,0.01830222209294637
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,65536,0.05045155684153239
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,16384,0.010399111443095738
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,51200,0.04188088907135857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,12288,0.008338666624493068
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,16384,0.02014133334159851
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,10240,0.007306666837798224
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,12288,0.01755822201569875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,8192,0.007848000360859765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,10240,0.01689511040846507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,7168,0.007340444458855524
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,8192,0.014888889259762235
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,6144,0.0069751110341813825
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,7168,0.015204444527626038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,5120,0.006576889091067844
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,6144,0.014231999715169271
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,4096,0.006945778098371293
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,5120,0.014144000079896716
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,3584,0.006619555668698416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,4096,0.01241422196229299
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,3584,0.013827555709415011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,3072,0.006305777778228124
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,2560,0.00628977765639623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,3072,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,2560,0.011759111450778114
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,2048,0.005922666854328579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,1536,0.004554666578769684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,2048,0.01108977778090371
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,1024,0.003917333152559069
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,1536,0.012062222593360476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,768,0.0035617777870761026
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,1024,0.010753778119881948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,512,0.0034871109657817413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,768,0.011745777395036487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,256,0.0028924445311228433
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,512,0.011067555182509951
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,128,0.0028737777223189673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,256,0.010748444332016839
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,64,0.0028568889117903183
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,128,0.009351111120647853
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,512,32,0.0025271111064487034
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,64,0.010128888818952773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,65536,0.014631110760900708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,512,32,0.009036444127559662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,51200,0.012855110896958245
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,65536,0.047188444270028
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,16384,0.008347555167145198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,51200,0.038739555411868624
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,12288,0.007982222570313348
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,16384,0.02018044392267863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,10240,0.007346666521496243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,12288,0.01752444439464145
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,8192,0.006620444357395172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,10240,0.016926222377353244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,7168,0.0063368889192740125
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,8192,0.014820444915029736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,6144,0.005941333456171884
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,7168,0.015221332510312399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,5120,0.005624889085690181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,6144,0.014120888378885059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,4096,0.005968889014588461
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,5120,0.014165333575672574
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,3584,0.005998222364319696
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,4096,0.012453333371215396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,3072,0.005635555419656966
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,3584,0.013471110827393003
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,2560,0.006046222315894232
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,3072,0.012071111136012606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,2048,0.005604444278611078
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,2560,0.012807110945383707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,2048,0.012181333369678922
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,1536,0.004286222159862518
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,1536,0.012086222569147745
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,1024,0.003710222327046924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,1024,0.011462222370836469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,768,0.003592888928121991
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,768,0.011791999969217511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,512,0.0032319999817344877
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,512,0.011797332929240333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,256,0.002946666752298673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,128,0.00286666676402092
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,256,0.010007999837398529
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,128,0.010116444693671333
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,64,0.002640888922744327
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,256,32,0.002918222298224767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,64,0.010058666268984476
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,65536,0.013348444468445249
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,256,32,0.00907733374171787
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,51200,0.01186222251918581
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,65536,0.04624266756905449
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,16384,0.007418666448858049
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,51200,0.038203554020987615
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,12288,0.007051555646790399
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,16384,0.019959999455346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,10240,0.006621333460013072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,12288,0.017709333035680983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,8192,0.0063662222690052455
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,10240,0.01680799987581041
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,7168,0.006320000108745363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,8192,0.015260444747077094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,6144,0.005719111197524601
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,7168,0.01517599986659156
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,6144,0.01423111061255137
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,5120,0.005960888746711943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,4096,0.005315555466545953
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,5120,0.013537777794731988
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,3584,0.00565422202150027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,4096,0.013433777623706393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,3072,0.005621333503060871
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,3584,0.012820444173283048
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,2560,0.005769777629110549
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,2560,0.012512000070677863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,3072,0.012783110969596438
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,2048,0.0052764444715446895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,1536,0.004223111189073986
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,2048,0.012128888732857175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,1024,0.003572444534964032
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,1536,0.01185955521133211
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,768,0.003540444291300244
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,1024,0.0114720000161065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,768,0.011114666859308878
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,512,0.0031973332580592898
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,256,0.0031511110977994073
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,512,0.01109955542617374
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,128,0.0028728888266616394
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,256,0.010759111079904767
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,64,0.002887111157178879
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,128,0.0090515555606948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,128,32,0.0025680000997251934
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,64,0.010009778042634329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,128,32,0.009012444151772393
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,65536,0.04585599899291992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,51200,0.009027555584907532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,51200,0.038155555725097656
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,16384,0.006608888920810487
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,16384,0.020042666130595736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,12288,0.005937777873542573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,65536,0.010465777582592435
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,12288,0.01761777698993683
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,10240,0.005649777750174205
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,8192,0.006303999986913469
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,10240,0.016197333733240765
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,7168,0.00565422202150027
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,8192,0.015533333023389181
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,6144,0.006036444256703059
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,7168,0.014836443795098199
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,5120,0.005657777604129579
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,6144,0.014507555299335055
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,4096,0.005607111172543631
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,5120,0.013776889277829064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,3584,0.005619555711746216
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,4096,0.013478221992651621
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,3072,0.005966222120655908
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,3584,0.013080000049538083
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,2560,0.005947555518812603
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,3072,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,2560,0.012525333298577202
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,2048,0.005671999934646819
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,1536,0.004232888834344016
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,2048,0.012070222033394707
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,1024,0.003846222327815162
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,1536,0.012064889073371887
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,768,0.0035031110876136353
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,1024,0.011433777709801992
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,512,0.0031964443624019623
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,768,0.01183733344078064
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,256,0.0028773333049482773
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,512,0.011358222199810876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,128,0.002874666617976295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,256,0.010743111371994019
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,128,0.010039110978444418
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,64,0.002575111058023241
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,64,0.010055999788973067
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,64,32,0.002567111204067866
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,64,32,0.010062221851613786
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,65536,0.0090515555606948
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,65536,0.04589066571659512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,51200,0.008024888734022776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,51200,0.03839822279082404
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,16384,0.0062871111763848206
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,16384,0.020091555184788175
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,12288,0.006312888943486744
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,10240,0.005931555396980709
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,12288,0.017080000705189176
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,8192,0.005233777893914117
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,10240,0.016527111331621807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,7168,0.005608888963858287
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,8192,0.014574221438831754
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,6144,0.005583999885453119
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,7168,0.014874666929244995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,5120,0.005568000177542369
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,6144,0.013774221969975365
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,4096,0.005280888742870755
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,5120,0.01407199932469262
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,4096,0.012122666670216454
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,3584,0.00591111100382275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,3072,0.005628444254398346
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,3584,0.01256977766752243
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,2560,0.005610666755172942
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,3072,0.012785777449607849
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,2048,0.005948444621430502
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,2560,0.011788444386588203
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,1536,0.004237333519591225
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,2048,0.012148444851239523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,1024,0.003882666842805015
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,1536,0.011358222199810876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,768,0.003216000066863166
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,1024,0.011456888582971362
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,512,0.0032684444967243406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,768,0.010823999842007955
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,256,0.00282577777074443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,512,0.011509333219793109
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,128,0.0028400001012616684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,256,0.009030222064918941
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,64,0.0024871110088295406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,128,0.010069333016872406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,64,0.007997333175606197
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,1,32,32,0.0027066665805048416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,51200,0.00045155557907289925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,fp8,1,32,32,0.008066666622956594
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,16384,0.0004817777727213171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,10240,0.00046933333699901897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,8192,0.00046666665002703667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,12288,0.0004746666592028406
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,7168,0.0004844444596932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,6144,0.0004293333428601424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,5120,0.0004320000039620532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,4096,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,3584,0.0003991111233416531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,3072,0.00046933333699901897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,2560,0.00040355554988814727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,2048,0.0004231110991289218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,1536,0.00043911109161045816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,1024,0.00045155557907289925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,768,0.00040088888878623646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,512,0.00040622221099005803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,256,0.0004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,128,0.0004275555515454875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,64,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,65536,32,0.00046577775436970923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,65536,0.00045600000561939343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,51200,0.00046933333699901897
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,16384,0.0004924444688691033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,12288,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,10240,0.0004622222234805425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,8192,0.00047377776354551315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,7168,0.0004817777727213171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,6144,0.00045333331864741113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,5120,0.00047377776354551315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,4096,0.00044711110078626213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,3584,0.00045155557907289925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,3072,0.0004604444321658876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,2560,0.0004844444596932994
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,2048,0.0004604444321658876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,1536,0.0004782222418321504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,1024,0.0004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,768,0.00045777779693404835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,512,0.0004702222326563464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,256,0.00047377776354551315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,128,0.000842666667368677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,64,0.0005146666533417172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,51200,32,0.00045600000561939343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,65536,0.0004728888678881857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,51200,0.0004782222418321504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,16384,0.001695999979144997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,12288,0.0004435555698970954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,10240,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,8192,0.0005946666416194703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,7168,0.00126933337499698
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,6144,0.00042577778610090416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,5120,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,4096,0.0004782222418321504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,3584,0.0004479999964435895
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,3072,0.0011555555587013562
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,2560,0.0004755555548601681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,2048,0.00046488891045252484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,1536,0.0005351110982398192
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,1024,0.00047644445051749546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,768,0.0009644444410999616
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,512,0.0004702222326563464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,256,0.0009902222082018852
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,128,0.0009555555880069733
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,64,0.000834666658192873
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,16384,32,0.0008053333084616396
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,65536,0.0008275555446743965
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,51200,0.000814222213294771
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,16384,0.0004924444688691033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,12288,0.000440888882925113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,10240,0.0004879999905824661
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,8192,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,7168,0.00045155557907289925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,6144,0.000842666667368677
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,5120,0.0004799999814066622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,4096,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,3584,0.00048266666837864456
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,3072,0.00045866664085123275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,2560,0.0004728888678881857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,2048,0.000606222233424584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,1536,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,1024,0.0004755555548601681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,768,0.00045866664085123275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,512,0.00041955556823975506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,256,0.0004888888862397935
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,128,0.0004702222326563464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,64,0.0004791110857493347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,12288,32,0.000793777768396669
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,65536,0.0007422222341928217
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,51200,0.0006888889086743196
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,16384,0.0006977777617673079
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,12288,0.0005848888928691546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,10240,0.0005831111015544997
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,8192,0.0004986666867302523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,7168,0.0005359999938971467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,6144,0.0005057778002487288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,5120,0.0005164444446563721
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,4096,0.000510222226795223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,3584,0.0005182222359710269
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,3072,0.0005422222117582957
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,2560,0.000544888898730278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,2048,0.0010319999936554167
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,1536,0.0004817777727213171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,1024,0.00042666665588816005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,768,0.0004604444321658876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,512,0.0004720000239710013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,256,0.00043555556072129146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,128,0.0004640000147951974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,64,0.00044622220512893464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,10240,32,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,65536,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,51200,0.0004373333520359463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,16384,0.0004231110991289218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,12288,0.0004364444563786189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,10240,0.000414222220165862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,8192,0.00041066666340662376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,7168,0.00045600000561939343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,6144,0.00039377777526776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,5120,0.00043822224769327376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,4096,0.0006213333561188645
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,3584,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,3072,0.0007991110906004906
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,2560,0.00041955556823975506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,2048,0.0004275555515454875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,1536,0.000420444438027011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,1024,0.00042666665588816005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,768,0.0006586666632857588
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,512,0.00041777777692510024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,256,0.00043377776940663654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,128,0.000414222220165862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,64,0.00041066666340662376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,8192,32,0.0006924444395634863
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,65536,0.0004542222143047386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,51200,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,16384,0.00042577778610090416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,12288,0.0004720000239710013
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,10240,0.0006577777676284313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,8192,0.00041955556823975506
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,7168,0.00045866664085123275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,6144,0.000510222226795223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,5120,0.0005724444571468566
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,4096,0.0007013333443966177
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,3584,0.0004951111041009426
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,3072,0.0005626666566563977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,2560,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,2048,0.0004302222126473983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,1536,0.0005742221967213684
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,1024,0.0004524444229900837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,768,0.00042488889044357673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,512,0.00043377776940663654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,256,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,128,0.0005546666474805938
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,64,0.00042666665588816005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,7168,32,0.0004942222084436152
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,65536,0.00040977776774929627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,51200,0.0003964444363696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,16384,0.0006239999913507038
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,12288,0.000440888882925113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,10240,0.0005084444354805682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,8192,0.0004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,7168,0.0005217777668601937
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,6144,0.00047377776354551315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,5120,0.0005564444387952486
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,4096,0.0005146666533417172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,3584,0.0004924444688691033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,3072,0.0005191111316283544
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,2560,0.0005208888712028662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,2048,0.0005173333403136995
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,1536,0.0004995555306474367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,1024,0.0005688888745175467
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,768,0.0004728888678881857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,512,0.0010364444719420539
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,256,0.0005244444538321759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,128,0.0005093333311378956
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,64,0.0005715555614895291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,6144,32,0.000544888898730278
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,65536,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,51200,0.0007271111114985413
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,16384,0.0005075555398232407
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,12288,0.0005404444204436408
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,10240,0.0005128889137672053
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,8192,0.0005839999972118272
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,7168,0.00046488891045252484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,6144,0.0005120000181098779
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,5120,0.0005946666416194703
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,4096,0.0005111111224525505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,3584,0.0005262222451468309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,3072,0.0004995555306474367
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,2560,0.0006222222000360489
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,2048,0.0006311111048691802
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,1536,0.00048533335535062686
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,1024,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,768,0.00046577775436970923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,512,0.0004373333520359463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,256,0.0004871110949251387
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,128,0.0004675555456843641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,64,0.0004755555548601681
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,5120,32,0.0004728888678881857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,65536,0.0004435555698970954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,51200,0.0004373333520359463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,16384,0.0004524444229900837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,12288,0.0004568889012767209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,10240,0.0004302222126473983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,8192,0.00043377776940663654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,7168,0.00043822224769327376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,6144,0.0006577777676284313
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,5120,0.00045777779693404835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,4096,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,3584,0.000440888882925113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,3072,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,2560,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,2048,0.0004799999814066622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,1536,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,1024,0.0004222222293416659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,768,0.0004302222126473983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,512,0.0004675555456843641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,256,0.0004364444563786189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,128,0.00047377776354551315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,64,0.00047644445051749546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,4096,32,0.0009066666372948223
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,65536,0.0006844444303876824
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,51200,0.00041244445472127857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,16384,0.0004231110991289218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,12288,0.0006302222092118528
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,10240,0.00040977776774929627
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,8192,0.00040088888878623646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,7168,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,6144,0.00039999999312890903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,5120,0.0003964444363696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,4096,0.0004160000114805168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,3584,0.00041511111582318944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,3072,0.0004924444688691033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,2560,0.00042577778610090416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,2048,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,1536,0.0003982222276843256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,1024,0.0004622222234805425
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,768,0.00043111110830472573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,512,0.0004373333520359463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,256,0.0003866666617492835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,128,0.000440888882925113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,64,0.0004435555698970954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3584,32,0.0005617777609990702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,65536,0.00043377776940663654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,51200,0.00041777777692510024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,16384,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,12288,0.00040355554988814727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,10240,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,8192,0.00041511111582318944
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,7168,0.00040800000230471295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,6144,0.00040800000230471295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,5120,0.00043466666506396397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,4096,0.00038577776609195606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,3584,0.000440888882925113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,3072,0.0004222222293416659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,2560,0.0004302222126473983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,2048,0.00043466666506396397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,1536,0.0007280000071558687
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,1024,0.00042666665588816005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,768,0.0005466666900449329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,512,0.0008248889094425571
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,256,0.00044177777858244046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,128,0.0006613333502577412
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,64,0.00043111110830472573
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,3072,32,0.0006533333410819372
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,65536,0.000414222220165862
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,51200,0.0005768888836933507
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,16384,0.00039377777526776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,12288,0.0009022222624884712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,10240,0.00037777778278622363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,8192,0.0004186666725824277
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,7168,0.0003964444363696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,6144,0.00040355554988814727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,5120,0.0004160000114805168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,4096,0.0004782222418321504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,3584,0.0004026666542308198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,3072,0.00040800000230471295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,2560,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,2048,0.00040800000230471295
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,1536,0.00040088888878623646
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,1024,0.00042488889044357673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,768,0.00040888889796204033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,512,0.00041244445472127857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,256,0.0004222222293416659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,128,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,64,0.00045777779693404835
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2560,32,0.0009022222624884712
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,65536,0.00043911109161045816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,51200,0.0003991111233416531
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,16384,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,12288,0.0004222222293416659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,10240,0.0004542222143047386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,8192,0.00040888889796204033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,7168,0.0004364444563786189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,6144,0.0004986666867302523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,5120,0.0005271110890640153
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,4096,0.00048355556403597194
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,3584,0.000648888914535443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,3072,0.0005057778002487288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,2560,0.0004782222418321504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,2048,0.0005342222025824918
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,1536,0.00047377776354551315
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,1024,0.0004799999814066622
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,768,0.0005057778002487288
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,512,0.0005031111132767465
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,256,0.0005111111224525505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,128,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,64,0.0004302222126473983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,2048,32,0.0005973333285914527
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,65536,0.0004231110991289218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,51200,0.00040622221099005803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,16384,0.0004320000039620532
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,12288,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,10240,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,8192,0.000428444447202815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,7168,0.00044711110078626213
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,6144,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,5120,0.0004791110857493347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,4096,0.00045066668341557187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,3584,0.0004817777727213171
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,3072,0.00049599999975827
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,2560,0.0004640000147951974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,2048,0.0004604444321658876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,1536,0.00040888889796204033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,1024,0.00045866664085123275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,768,0.00045066668341557187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,512,0.0005911111107303037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,256,0.0004542222143047386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,128,0.0007564444612297746
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,64,0.000655111132396592
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1536,32,0.000379555548230807
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,65536,0.0006248888870080312
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,51200,0.0005626666566563977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,16384,0.0005715555614895291
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,12288,0.0005466666900449329
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,10240,0.000504000008934074
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,8192,0.0005111111224525505
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,7168,0.00037777778278622363
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,6144,0.0004791110857493347
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,5120,0.0005262222451468309
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,4096,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,3584,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,3072,0.0005146666533417172
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,2560,0.00042488889044357673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,2048,0.00045600000561939343
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,1536,0.0004275555515454875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,1024,0.0004293333428601424
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,768,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,512,0.00043999998726778564
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,256,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,128,0.00042488889044357673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,64,0.0004373333520359463
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,1024,32,0.00041066666340662376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,65536,0.0004213333336843385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,51200,0.00039466667092508736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,16384,0.0003982222276843256
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,12288,0.0004026666542308198
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,10240,0.0003893333228511943
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,8192,0.00041244445472127857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,7168,0.0004542222143047386
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,6144,0.0004053333412028021
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,5120,0.00039999999312890903
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,4096,0.00041777777692510024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,3584,0.0005084444354805682
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,3072,0.0004222222293416659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,2560,0.0004302222126473983
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,2048,0.00046577775436970923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,1536,0.00043911109161045816
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,1024,0.0004986666867302523
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,768,0.00041244445472127857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,512,0.0004924444688691033
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,256,0.0004640000147951974
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,128,0.00046666665002703667
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,64,0.00045066668341557187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,768,32,0.00045333331864741113
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,65536,0.0005626666566563977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,51200,0.000420444438027011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,16384,0.00042577778610090416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,12288,0.0004524444229900837
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,10240,0.0004275555515454875
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,8192,0.00043466666506396397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,7168,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,6144,0.00039555556658241485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,5120,0.00045155557907289925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,4096,0.00046577775436970923
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,3584,0.0005297777760359977
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,3072,0.00043555556072129146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,2560,0.0004231110991289218
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,2048,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,1536,0.0007635555747482511
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,1024,0.0004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,768,0.00040622221099005803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,512,0.0004568889012767209
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,256,0.0004222222293416659
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,128,0.0010195555579331187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,64,0.0004728888678881857
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,512,32,0.0005537778035634094
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,65536,0.0005324444630079799
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,51200,0.00039377777526776
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,16384,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,12288,0.0004364444563786189
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,10240,0.00043822224769327376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,8192,0.00042577778610090416
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,7168,0.0004160000114805168
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,6144,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,5120,0.000428444447202815
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,4096,0.00043377776940663654
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,3584,0.00046488891045252484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,3072,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,2560,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,2048,0.00045066668341557187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,1536,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,1024,0.00042399999478624924
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,768,0.00041777777692510024
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,512,0.00044888889210091694
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,256,0.00044177777858244046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,128,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,64,0.0004435555698970954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,256,32,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,65536,0.0003973333320269982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,51200,0.0004711111283136739
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,16384,0.00041688888126777276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,12288,0.00041066666340662376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,10240,0.00043555556072129146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,8192,0.00044622220512893464
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,7168,0.00043288889961938065
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,6144,0.00045066668341557187
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,5120,0.0004675555456843641
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,4096,0.00041688888126777276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,3584,0.00044444446555442276
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,3072,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,2560,0.00040622221099005803
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,2048,0.0008071110997762945
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,1536,0.0007360000163316727
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,1024,0.000606222233424584
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,768,0.0006257777826653586
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,512,0.0005911111107303037
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,256,0.0005582222301099036
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,128,0.0005617777609990702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,64,0.0005244444538321759
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,128,32,0.0005208888712028662
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,65536,0.0004782222418321504
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,51200,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,16384,0.0005288888803786702
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,12288,0.00047644445051749546
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,10240,0.00044977778775824443
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,8192,0.00043555556072129146
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,7168,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,6144,0.00042488889044357673
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,5120,0.00043466666506396397
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,4096,0.0004604444321658876
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,3584,0.00045866664085123275
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,3072,0.0004213333336843385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,2560,0.0003724444347123305
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,2048,0.0004115555590639512
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,1536,0.00044266667423976795
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,1024,0.00042666665588816005
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,768,0.0004453333094716072
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,512,0.00044177777858244046
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,256,0.0004213333336843385
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,128,0.00039022221850852173
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,64,0.0005022222176194191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,64,32,0.0008462221982578436
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,65536,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,51200,0.00038577776609195606
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,16384,0.000420444438027011
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,12288,0.00041066666340662376
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,10240,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,8192,0.00038400000064737263
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,7168,0.0003920000098231766
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,6144,0.0003973333320269982
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,5120,0.0004435555698970954
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,4096,0.00039555556658241485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,3584,0.0004133333245085345
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,3072,0.0003964444363696708
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,2560,0.00039555556658241485
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,2048,0.00039466667092508736
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,1536,0.0003911111141658491
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,1024,0.0005022222176194191
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,768,0.0004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,512,0.00046488891045252484
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,256,0.0004613333278232151
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,128,0.00045155557907289925
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,64,0.0004684444413416916
VLLM,0.14.1.dev1+gd68209402,NVIDIA GB300,gemm,vllm_default,float16,0,32,32,0.00041511111582318944
